M.A. Migut
info
Please Note
<p>This page displays the records of the person named above and is not linked to a unique person identifier. This record may need to be merged to a profile.</p>
53 records found
1
Small drones increasingly operate indoors, where GPS is unavailable and the heavier sensors normally used for indoor positioning exceed a micro-drone's weight, size, and power budget. Visible light is a practical alternative, because the spaces a drone operates in are already lit. An ordinary lamp can be modulated to act as a navigation beacon, while the drone carries only a small light sensor to detect it. Prior work by Huang et al. navigates a micro-drone toward a single such beacon, recovering the direction of the light from a ring of photodiodes, but it keeps no record of the light field. It therefore cannot distinguish one beacon from another, nor route around an obstacle that blocks the light.
This thesis gives a weight-, size-, and power-constrained micro-drone the navigation those methods lacked, using only on-board light sensing. We design a fused controller that combines two directions. The first is the bearing toward the beacon, read from a ring of photodiodes. The second is a gradient, which the drone estimates from a map of the light field that it builds as it flies. The sensing runs on a custom light-sensing board that fits well within a micro-drone's payload budget. A signal pipeline on the drone then recovers each beacon from the board's noisy output through a frequency analysis, and reports a confidence measure for each reading. We validate the approach in a Webots simulation and in real flight. The drone reaches modulated beacons forming a predetermined path, and routes around an obstacle, using its stored map of the light field when the obstacle blocks the beacon. ...
This thesis gives a weight-, size-, and power-constrained micro-drone the navigation those methods lacked, using only on-board light sensing. We design a fused controller that combines two directions. The first is the bearing toward the beacon, read from a ring of photodiodes. The second is a gradient, which the drone estimates from a map of the light field that it builds as it flies. The sensing runs on a custom light-sensing board that fits well within a micro-drone's payload budget. A signal pipeline on the drone then recovers each beacon from the board's noisy output through a frequency analysis, and reports a confidence measure for each reading. We validate the approach in a Webots simulation and in real flight. The drone reaches modulated beacons forming a predetermined path, and routes around an obstacle, using its stored map of the light field when the obstacle blocks the beacon. ...
Small drones increasingly operate indoors, where GPS is unavailable and the heavier sensors normally used for indoor positioning exceed a micro-drone's weight, size, and power budget. Visible light is a practical alternative, because the spaces a drone operates in are already lit. An ordinary lamp can be modulated to act as a navigation beacon, while the drone carries only a small light sensor to detect it. Prior work by Huang et al. navigates a micro-drone toward a single such beacon, recovering the direction of the light from a ring of photodiodes, but it keeps no record of the light field. It therefore cannot distinguish one beacon from another, nor route around an obstacle that blocks the light.
This thesis gives a weight-, size-, and power-constrained micro-drone the navigation those methods lacked, using only on-board light sensing. We design a fused controller that combines two directions. The first is the bearing toward the beacon, read from a ring of photodiodes. The second is a gradient, which the drone estimates from a map of the light field that it builds as it flies. The sensing runs on a custom light-sensing board that fits well within a micro-drone's payload budget. A signal pipeline on the drone then recovers each beacon from the board's noisy output through a frequency analysis, and reports a confidence measure for each reading. We validate the approach in a Webots simulation and in real flight. The drone reaches modulated beacons forming a predetermined path, and routes around an obstacle, using its stored map of the light field when the obstacle blocks the beacon.
This thesis gives a weight-, size-, and power-constrained micro-drone the navigation those methods lacked, using only on-board light sensing. We design a fused controller that combines two directions. The first is the bearing toward the beacon, read from a ring of photodiodes. The second is a gradient, which the drone estimates from a map of the light field that it builds as it flies. The sensing runs on a custom light-sensing board that fits well within a micro-drone's payload budget. A signal pipeline on the drone then recovers each beacon from the board's noisy output through a frequency analysis, and reports a confidence measure for each reading. We validate the approach in a Webots simulation and in real flight. The drone reaches modulated beacons forming a predetermined path, and routes around an obstacle, using its stored map of the light field when the obstacle blocks the beacon.
Good Enough to Talk To?
Perceived Usefulness and Social Confidence development in a GenAI Simulated Team Interaction Study
This research investigates the consequences for negotiation and communication skills perceived by students after using a generative AI chatbot to practice human-centered activities within the field of aerospace engineering, as well as the pedagogical usefulness of such a tool. Generative Artificial Intelligence (GenAI) is increasingly integrated in engineering education, offering new ways to simulate job-related activities and give students more opportunities for practice. However, most existing research focuses on technical performance, such as coding support and general usability, while the effects on social and communication skills remain less explored.
Seventeen aerospace engineering students from TU Delft participated in a negotiation scenario with a ChatGPT-based chatbot acting as a professional teammate, followed by a mixed questionnaire. Results show that students generally perceive the chatbot as easy to use (89% positive), while holding a neutral-positive attitude towards its performance (55.29% positive) and professional relevance (68% positive). Qualitative findings suggest that the chatbot is mainly valued for preparing and structuring information, but not for improving communication skills. This is reflected in mixed future-use intentions (43.8% positive and negative responses; M = -0.31). While perceived social confidence remained relatively high (79.4% positive), no clear improvement in communication ability was reported. The chatbot interactions were also consistently viewed as less realistic than human interactions due to the lack of emotional expression and non-verbal means of communication.
Overall, the chatbot is viewed as a useful tool for preparation, but not as a replacement for human-centered communication training. These findings add to the understanding of how GenAI chatbots function in educational settings and offer initial considerations for their responsible use in aerospace engineering education. ...
Seventeen aerospace engineering students from TU Delft participated in a negotiation scenario with a ChatGPT-based chatbot acting as a professional teammate, followed by a mixed questionnaire. Results show that students generally perceive the chatbot as easy to use (89% positive), while holding a neutral-positive attitude towards its performance (55.29% positive) and professional relevance (68% positive). Qualitative findings suggest that the chatbot is mainly valued for preparing and structuring information, but not for improving communication skills. This is reflected in mixed future-use intentions (43.8% positive and negative responses; M = -0.31). While perceived social confidence remained relatively high (79.4% positive), no clear improvement in communication ability was reported. The chatbot interactions were also consistently viewed as less realistic than human interactions due to the lack of emotional expression and non-verbal means of communication.
Overall, the chatbot is viewed as a useful tool for preparation, but not as a replacement for human-centered communication training. These findings add to the understanding of how GenAI chatbots function in educational settings and offer initial considerations for their responsible use in aerospace engineering education. ...
This research investigates the consequences for negotiation and communication skills perceived by students after using a generative AI chatbot to practice human-centered activities within the field of aerospace engineering, as well as the pedagogical usefulness of such a tool. Generative Artificial Intelligence (GenAI) is increasingly integrated in engineering education, offering new ways to simulate job-related activities and give students more opportunities for practice. However, most existing research focuses on technical performance, such as coding support and general usability, while the effects on social and communication skills remain less explored.
Seventeen aerospace engineering students from TU Delft participated in a negotiation scenario with a ChatGPT-based chatbot acting as a professional teammate, followed by a mixed questionnaire. Results show that students generally perceive the chatbot as easy to use (89% positive), while holding a neutral-positive attitude towards its performance (55.29% positive) and professional relevance (68% positive). Qualitative findings suggest that the chatbot is mainly valued for preparing and structuring information, but not for improving communication skills. This is reflected in mixed future-use intentions (43.8% positive and negative responses; M = -0.31). While perceived social confidence remained relatively high (79.4% positive), no clear improvement in communication ability was reported. The chatbot interactions were also consistently viewed as less realistic than human interactions due to the lack of emotional expression and non-verbal means of communication.
Overall, the chatbot is viewed as a useful tool for preparation, but not as a replacement for human-centered communication training. These findings add to the understanding of how GenAI chatbots function in educational settings and offer initial considerations for their responsible use in aerospace engineering education.
Seventeen aerospace engineering students from TU Delft participated in a negotiation scenario with a ChatGPT-based chatbot acting as a professional teammate, followed by a mixed questionnaire. Results show that students generally perceive the chatbot as easy to use (89% positive), while holding a neutral-positive attitude towards its performance (55.29% positive) and professional relevance (68% positive). Qualitative findings suggest that the chatbot is mainly valued for preparing and structuring information, but not for improving communication skills. This is reflected in mixed future-use intentions (43.8% positive and negative responses; M = -0.31). While perceived social confidence remained relatively high (79.4% positive), no clear improvement in communication ability was reported. The chatbot interactions were also consistently viewed as less realistic than human interactions due to the lack of emotional expression and non-verbal means of communication.
Overall, the chatbot is viewed as a useful tool for preparation, but not as a replacement for human-centered communication training. These findings add to the understanding of how GenAI chatbots function in educational settings and offer initial considerations for their responsible use in aerospace engineering education.
Does Representation Matter? Comparing Algebraic and Geometric Approaches to Teaching L1/L2 Regularization
Effects on Conceptual Understanding, Problem-Solving, and Knowledge Transfer
As machine learning becomes a standard part of computing and engineering curricula, teaching its core concepts effectively has become an important educational challenge. Regularization is one such concept: L1 and L2 penalties are widely used to prevent overfitting, with L1 producing sparse solutions and L2 shrinking weights more smoothly. It is commonly taught through two representational formats: an algebraic one, presenting the loss function and its penalty terms, and a geometric one, depicting constraint regions and their intersection with the loss contours. However, instructors choose between them without clear evidence on which better supports learning. This study asks whether teaching regularization algebraically or geometrically leads to different student performance, and in which kinds of understanding.
Two interactive notebooks, matched on learning objectives, length, and reading difficulty, were developed to teach the concept in each format and are released openly; they were compared in a between-subjects experiment with students who had completed an introductory machine learning course. Learning was measured with a post-test spanning conceptual understanding, problem-solving, and knowledge transfer, alongside a thematic analysis of students’ written explanations.
Both formats supported practical reasoning about regularization, but not identically: the algebraic group performed better overall, with its clearest advantage in conceptual understanding, no reliable difference on problem-solving, and an inconclusive result on transfer. The two groups largely shared the same core understanding but expressed it through different vocabularies: a penalty on the loss function versus a shrinking constraint region in weight space. Representational choice therefore appears to shape how students explain regularization more than whether they grasp it, suggesting that combining the two formats may best support learning.
...
Two interactive notebooks, matched on learning objectives, length, and reading difficulty, were developed to teach the concept in each format and are released openly; they were compared in a between-subjects experiment with students who had completed an introductory machine learning course. Learning was measured with a post-test spanning conceptual understanding, problem-solving, and knowledge transfer, alongside a thematic analysis of students’ written explanations.
Both formats supported practical reasoning about regularization, but not identically: the algebraic group performed better overall, with its clearest advantage in conceptual understanding, no reliable difference on problem-solving, and an inconclusive result on transfer. The two groups largely shared the same core understanding but expressed it through different vocabularies: a penalty on the loss function versus a shrinking constraint region in weight space. Representational choice therefore appears to shape how students explain regularization more than whether they grasp it, suggesting that combining the two formats may best support learning.
...
As machine learning becomes a standard part of computing and engineering curricula, teaching its core concepts effectively has become an important educational challenge. Regularization is one such concept: L1 and L2 penalties are widely used to prevent overfitting, with L1 producing sparse solutions and L2 shrinking weights more smoothly. It is commonly taught through two representational formats: an algebraic one, presenting the loss function and its penalty terms, and a geometric one, depicting constraint regions and their intersection with the loss contours. However, instructors choose between them without clear evidence on which better supports learning. This study asks whether teaching regularization algebraically or geometrically leads to different student performance, and in which kinds of understanding.
Two interactive notebooks, matched on learning objectives, length, and reading difficulty, were developed to teach the concept in each format and are released openly; they were compared in a between-subjects experiment with students who had completed an introductory machine learning course. Learning was measured with a post-test spanning conceptual understanding, problem-solving, and knowledge transfer, alongside a thematic analysis of students’ written explanations.
Both formats supported practical reasoning about regularization, but not identically: the algebraic group performed better overall, with its clearest advantage in conceptual understanding, no reliable difference on problem-solving, and an inconclusive result on transfer. The two groups largely shared the same core understanding but expressed it through different vocabularies: a penalty on the loss function versus a shrinking constraint region in weight space. Representational choice therefore appears to shape how students explain regularization more than whether they grasp it, suggesting that combining the two formats may best support learning.
Two interactive notebooks, matched on learning objectives, length, and reading difficulty, were developed to teach the concept in each format and are released openly; they were compared in a between-subjects experiment with students who had completed an introductory machine learning course. Learning was measured with a post-test spanning conceptual understanding, problem-solving, and knowledge transfer, alongside a thematic analysis of students’ written explanations.
Both formats supported practical reasoning about regularization, but not identically: the algebraic group performed better overall, with its clearest advantage in conceptual understanding, no reliable difference on problem-solving, and an inconclusive result on transfer. The two groups largely shared the same core understanding but expressed it through different vocabularies: a penalty on the loss function versus a shrinking constraint region in weight space. Representational choice therefore appears to shape how students explain regularization more than whether they grasp it, suggesting that combining the two formats may best support learning.
Good Enough to Talk To?
Exploring the Acceptance and Social Presence of AI Chatbots for Human-Centered Task Training in Electrical Engineering Education
Generative AI chatbots are increasingly proposed as simulated stakeholders. These stakeholders let engineering students rehearse human-centered communication without the scheduling and access constraints of real interlocutors. In an engineering education context, how students receive such tools, and whether they engage with them as social partners, remains little studied. This paper examines the acceptance and perceived social presence of an AI-simulated non-technical stakeholder in an electrical engineering setting. Fifteen students each completed a single customer discovery interview with a large language model chatbot, framed around the Value Proposition Canvas. The study follows a convergent mixed-methods design, pairing UTAUT-derived acceptance constructs and an adapted social presence scale with a reflexive thematic analysis of the interaction transcripts and written reflections. Acceptance was high but conditional, and tracked how useful students judged the tool rather than how easy it was to use. Social presence was the lowest and most variable of the measured constructs and moved largely independently of perceived usefulness, so a more socially present chatbot did not register as a more useful one. Students valued the tool as an accessible option when real stakeholders were hard to reach, while noting absent empathy, predictable responses, and a risk of over-reliance. Transcripts show some students got the responses they wanted by engineering their prompts rather than by communicating as they would with a real stakeholder. These self-reported patterns indicate that such chatbots are best positioned as carefully bounded preparation before real stakeholder contact rather than as replacements for it.
...
Generative AI chatbots are increasingly proposed as simulated stakeholders. These stakeholders let engineering students rehearse human-centered communication without the scheduling and access constraints of real interlocutors. In an engineering education context, how students receive such tools, and whether they engage with them as social partners, remains little studied. This paper examines the acceptance and perceived social presence of an AI-simulated non-technical stakeholder in an electrical engineering setting. Fifteen students each completed a single customer discovery interview with a large language model chatbot, framed around the Value Proposition Canvas. The study follows a convergent mixed-methods design, pairing UTAUT-derived acceptance constructs and an adapted social presence scale with a reflexive thematic analysis of the interaction transcripts and written reflections. Acceptance was high but conditional, and tracked how useful students judged the tool rather than how easy it was to use. Social presence was the lowest and most variable of the measured constructs and moved largely independently of perceived usefulness, so a more socially present chatbot did not register as a more useful one. Students valued the tool as an accessible option when real stakeholders were hard to reach, while noting absent empathy, predictable responses, and a risk of over-reliance. Transcripts show some students got the responses they wanted by engineering their prompts rather than by communicating as they would with a real stakeholder. These self-reported patterns indicate that such chatbots are best positioned as carefully bounded preparation before real stakeholder contact rather than as replacements for it.
Simulating Stakeholders: Generative AI Chatbots in Architecture Education
Perceived Usefulness and Diversity Awareness in a Simulated Interview Study
Generative AI chatbots are increasingly used in engineering education to simulate human stakeholder interactions at scale. While existing studies demonstrate their potential for hands-on learning, the consequences of replacing real human interaction with AI simulation remain largely unexplored. This study investigates the implications of using generative AI chatbots for simulated human interaction on architecture students' awareness of human diversity. A ChatGPT-based chatbot simulating a secondary school teacher in Rotterdam was developed and evaluated with 14 architecture students from TU Delft. Participants completed a 25-minute simulated user interview followed by a mixed-methods evaluation consisting of validated questionnaire scales and open-ended reflection questions. Quantitative data were analyzed using descriptive statistics and qualitative data were analyzed using thematic analysis. Results indicate that students rated the chatbot highly on ease of use but showed more moderate scores on perceived usefulness and professional relevance. Physical mobility dominated diversity consideration (93%), while cultural background and visual impairment were rarely considered (36% and 14% respectively). Qualitative themes suggest that the chatbot functioned as a gap-filler in architecture education and prompted students to consider overlooked user groups, but that this awareness was largely chatbot-driven rather than self-initiated. Students consistently positioned the tool as a useful supplement for study contexts rather than a substitute for real human interaction. These findings contribute to a growing understanding of the pedagogical implications of AI chatbots in human-centered education, and offer preliminary guidance for their responsible use in architecture curricula.
...
Generative AI chatbots are increasingly used in engineering education to simulate human stakeholder interactions at scale. While existing studies demonstrate their potential for hands-on learning, the consequences of replacing real human interaction with AI simulation remain largely unexplored. This study investigates the implications of using generative AI chatbots for simulated human interaction on architecture students' awareness of human diversity. A ChatGPT-based chatbot simulating a secondary school teacher in Rotterdam was developed and evaluated with 14 architecture students from TU Delft. Participants completed a 25-minute simulated user interview followed by a mixed-methods evaluation consisting of validated questionnaire scales and open-ended reflection questions. Quantitative data were analyzed using descriptive statistics and qualitative data were analyzed using thematic analysis. Results indicate that students rated the chatbot highly on ease of use but showed more moderate scores on perceived usefulness and professional relevance. Physical mobility dominated diversity consideration (93%), while cultural background and visual impairment were rarely considered (36% and 14% respectively). Qualitative themes suggest that the chatbot functioned as a gap-filler in architecture education and prompted students to consider overlooked user groups, but that this awareness was largely chatbot-driven rather than self-initiated. Students consistently positioned the tool as a useful supplement for study contexts rather than a substitute for real human interaction. These findings contribute to a growing understanding of the pedagogical implications of AI chatbots in human-centered education, and offer preliminary guidance for their responsible use in architecture curricula.
The emergence of generative Artificial Intelligence (AI) chatbots allows engineering, and design programs specifically, to simulate interactions with humans at low cost, but little research has been done on students' acceptance of and perceived empathic connection to AI chatbots. This paper conducts a user study to evaluate aforementioned aspects when practicing design interviews with students from the Industrial Design Engineering faculty at Delft University of Technology. Nine participants completed an exercise involving interviewing an AI persona and answered a questionnaire that adapted subscales from UTAUT2, COLLES, and EMPA-D. The students rated the chatbot consistently positive on ease of use and judged it moderately useful and professionally relevant, though opinions on its usefulness and their intention to keep using it varied widely. Self-reported empathy toward the persona was high on perspective-taking and self-awareness but lower and more variable on shared personal experience. These exploratory findings suggest that similar chatbots can provide accessible, low-stakes practice that engages students' empathic perspective-taking, while adoption and affective connection with a dissimilar persona remain open challenges. For educators in design programs, AI chatbots appear viable for low-stakes interview training, yet careful persona design and strategies for sustained engagement are needed before broader adoption.
...
The emergence of generative Artificial Intelligence (AI) chatbots allows engineering, and design programs specifically, to simulate interactions with humans at low cost, but little research has been done on students' acceptance of and perceived empathic connection to AI chatbots. This paper conducts a user study to evaluate aforementioned aspects when practicing design interviews with students from the Industrial Design Engineering faculty at Delft University of Technology. Nine participants completed an exercise involving interviewing an AI persona and answered a questionnaire that adapted subscales from UTAUT2, COLLES, and EMPA-D. The students rated the chatbot consistently positive on ease of use and judged it moderately useful and professionally relevant, though opinions on its usefulness and their intention to keep using it varied widely. Self-reported empathy toward the persona was high on perspective-taking and self-awareness but lower and more variable on shared personal experience. These exploratory findings suggest that similar chatbots can provide accessible, low-stakes practice that engages students' empathic perspective-taking, while adoption and affective connection with a dissimilar persona remain open challenges. For educators in design programs, AI chatbots appear viable for low-stakes interview training, yet careful persona design and strategies for sustained engagement are needed before broader adoption.
Teaching Principal Component Analysis Through Multiple Representations
Impacts on Conceptual Understanding, Problem Solving and Knowledge Transfer
Machine Learning has become a prominent component of STEM curricula, yet there remains a lack of standardized pedagogical frameworks for teaching it. Principal Component Analysis represents a typical educational challenge, requiring students to simultaneously coordinate algebraic manipulation, geometric intuition, and algorithmic thinking. This study evaluates how different combinations of instructional representations affect undergraduate students’ conceptual understanding, problem-solving performance, and knowledge transfer ability. An experiment was conducted with 27 first-year Computer Science students at TU Delft. Participants were assigned to one of two experimental conditions: a traditional static representations group or a multimedia-enhanced representations group. Quantitative analysis revealed that the static group significantly outperformed the interactive group in total post-test scores and knowledge transfer. No statistically significant differences were observed in conceptual understanding or problem-solving performance. Qualitative thematic analysis indicated a disconnect between perceived and actual learning: while students preferred the interactive widgets for building geometric intuition, these features may have introduced extraneous cognitive load, provided a false sense of understanding, or students simply ran out of time because interactive exploration takes longer.
...
Machine Learning has become a prominent component of STEM curricula, yet there remains a lack of standardized pedagogical frameworks for teaching it. Principal Component Analysis represents a typical educational challenge, requiring students to simultaneously coordinate algebraic manipulation, geometric intuition, and algorithmic thinking. This study evaluates how different combinations of instructional representations affect undergraduate students’ conceptual understanding, problem-solving performance, and knowledge transfer ability. An experiment was conducted with 27 first-year Computer Science students at TU Delft. Participants were assigned to one of two experimental conditions: a traditional static representations group or a multimedia-enhanced representations group. Quantitative analysis revealed that the static group significantly outperformed the interactive group in total post-test scores and knowledge transfer. No statistically significant differences were observed in conceptual understanding or problem-solving performance. Qualitative thematic analysis indicated a disconnect between perceived and actual learning: while students preferred the interactive widgets for building geometric intuition, these features may have introduced extraneous cognitive load, provided a false sense of understanding, or students simply ran out of time because interactive exploration takes longer.
Teaching Gradient Descent
An Exploratory Study on Classic Textbook vs. Multiple Representations Approaches
Machine learning is increasingly important in computer science education, but introductory concepts can be difficult because they combine mathematical notation, algorithmic reasoning, and conceptual understanding. Gradient descent is one such concept: students may reproduce the update rule while still struggling to explain the role of the loss function, gradient, learning rate, and repeated parameter updates.
This paper investigates whether multiple representations can support beginner understanding compared with a classic textbook-style explanation.
A small-scale exploratory experiment was conducted with students who had little or no prior machine learning experience. Participants completed a prerequisite pre-test, studied gradient descent using either a text-and-formula-based explanation or a multiple-representations explanation, completed a post-test, and answered an experience survey.
The multiple-representations condition showed higher post-test performance, especially on computation and application tasks, as well as higher confidence, clarity, usefulness, and engagement. Perceived cognitive load remained similar across conditions.
These findings suggest that aligned multiple representations can help beginners connect formal notation with concrete calculations and intuitive understanding, although the results should be interpreted cautiously because of the small sample size. ...
This paper investigates whether multiple representations can support beginner understanding compared with a classic textbook-style explanation.
A small-scale exploratory experiment was conducted with students who had little or no prior machine learning experience. Participants completed a prerequisite pre-test, studied gradient descent using either a text-and-formula-based explanation or a multiple-representations explanation, completed a post-test, and answered an experience survey.
The multiple-representations condition showed higher post-test performance, especially on computation and application tasks, as well as higher confidence, clarity, usefulness, and engagement. Perceived cognitive load remained similar across conditions.
These findings suggest that aligned multiple representations can help beginners connect formal notation with concrete calculations and intuitive understanding, although the results should be interpreted cautiously because of the small sample size. ...
Machine learning is increasingly important in computer science education, but introductory concepts can be difficult because they combine mathematical notation, algorithmic reasoning, and conceptual understanding. Gradient descent is one such concept: students may reproduce the update rule while still struggling to explain the role of the loss function, gradient, learning rate, and repeated parameter updates.
This paper investigates whether multiple representations can support beginner understanding compared with a classic textbook-style explanation.
A small-scale exploratory experiment was conducted with students who had little or no prior machine learning experience. Participants completed a prerequisite pre-test, studied gradient descent using either a text-and-formula-based explanation or a multiple-representations explanation, completed a post-test, and answered an experience survey.
The multiple-representations condition showed higher post-test performance, especially on computation and application tasks, as well as higher confidence, clarity, usefulness, and engagement. Perceived cognitive load remained similar across conditions.
These findings suggest that aligned multiple representations can help beginners connect formal notation with concrete calculations and intuitive understanding, although the results should be interpreted cautiously because of the small sample size.
This paper investigates whether multiple representations can support beginner understanding compared with a classic textbook-style explanation.
A small-scale exploratory experiment was conducted with students who had little or no prior machine learning experience. Participants completed a prerequisite pre-test, studied gradient descent using either a text-and-formula-based explanation or a multiple-representations explanation, completed a post-test, and answered an experience survey.
The multiple-representations condition showed higher post-test performance, especially on computation and application tasks, as well as higher confidence, clarity, usefulness, and engagement. Perceived cognitive load remained similar across conditions.
These findings suggest that aligned multiple representations can help beginners connect formal notation with concrete calculations and intuitive understanding, although the results should be interpreted cautiously because of the small sample size.
Teaching Decision Trees in Machine Learning using multiple representations
Effects on Conceptual understanding, Problem-solving performance, and Knowledge transfer ability
Bachelor thesis
(2026)
-
C.K.O. Tran Chau Kieu Oanh, I.E.I. Rențea, M.A. Migut, Jorge Abraham Martinez Castaneda
Machine Learning (ML) is a rapidly growing field within Artificial Intelligence and one of the most prominent areas of technological study, which is particularly challenging for new learners as it requires a strong grasp of abstract algorithmic structures along with rigorous mathematical reasoning. Despite this, traditional instructional approaches often fail to support deep conceptual understanding, particularly for foundational models such as Decision Trees. This study examines whether integrating multiple instructional representations (including textual explanations, visualizations, analogies, videos, and interactive simulations) enhances student learning outcomes compared to traditional text-only materials.
To examine the lack of empirical evidence on multi-representational teaching for Decision Trees in ML education, a mixed-methods pilot experiment was employed with 10 participants was employed, comparing a multi-representation tutorial group to a text-only group. After a pre-test on mathematical and logical reasoning, participants completed a structured learning phase and a post-test measuring conceptual understanding, problem-solving, and transfer. Semi-structured interviews were also conducted to capture learner experiences.
Quantitative analysis included independent group comparisons using descriptive statistics, and inferential tests. Qualitative data were analyzed with inductive and deductive thematic analysis.
This study provides preliminary evidence that multi-representational instruction may improve problem-solving performance in Decision Tree learning. However, due to the small sample size and lack of statistical significance on two of three outcomes, these findings should be interpreted cautiously and require replication with larger samples.
The study contributes a structured evaluation framework for multi-representational ML education and provides evidence supporting the potential benefits of interactive instructional design in teaching Decision Trees. ...
To examine the lack of empirical evidence on multi-representational teaching for Decision Trees in ML education, a mixed-methods pilot experiment was employed with 10 participants was employed, comparing a multi-representation tutorial group to a text-only group. After a pre-test on mathematical and logical reasoning, participants completed a structured learning phase and a post-test measuring conceptual understanding, problem-solving, and transfer. Semi-structured interviews were also conducted to capture learner experiences.
Quantitative analysis included independent group comparisons using descriptive statistics, and inferential tests. Qualitative data were analyzed with inductive and deductive thematic analysis.
This study provides preliminary evidence that multi-representational instruction may improve problem-solving performance in Decision Tree learning. However, due to the small sample size and lack of statistical significance on two of three outcomes, these findings should be interpreted cautiously and require replication with larger samples.
The study contributes a structured evaluation framework for multi-representational ML education and provides evidence supporting the potential benefits of interactive instructional design in teaching Decision Trees. ...
Machine Learning (ML) is a rapidly growing field within Artificial Intelligence and one of the most prominent areas of technological study, which is particularly challenging for new learners as it requires a strong grasp of abstract algorithmic structures along with rigorous mathematical reasoning. Despite this, traditional instructional approaches often fail to support deep conceptual understanding, particularly for foundational models such as Decision Trees. This study examines whether integrating multiple instructional representations (including textual explanations, visualizations, analogies, videos, and interactive simulations) enhances student learning outcomes compared to traditional text-only materials.
To examine the lack of empirical evidence on multi-representational teaching for Decision Trees in ML education, a mixed-methods pilot experiment was employed with 10 participants was employed, comparing a multi-representation tutorial group to a text-only group. After a pre-test on mathematical and logical reasoning, participants completed a structured learning phase and a post-test measuring conceptual understanding, problem-solving, and transfer. Semi-structured interviews were also conducted to capture learner experiences.
Quantitative analysis included independent group comparisons using descriptive statistics, and inferential tests. Qualitative data were analyzed with inductive and deductive thematic analysis.
This study provides preliminary evidence that multi-representational instruction may improve problem-solving performance in Decision Tree learning. However, due to the small sample size and lack of statistical significance on two of three outcomes, these findings should be interpreted cautiously and require replication with larger samples.
The study contributes a structured evaluation framework for multi-representational ML education and provides evidence supporting the potential benefits of interactive instructional design in teaching Decision Trees.
To examine the lack of empirical evidence on multi-representational teaching for Decision Trees in ML education, a mixed-methods pilot experiment was employed with 10 participants was employed, comparing a multi-representation tutorial group to a text-only group. After a pre-test on mathematical and logical reasoning, participants completed a structured learning phase and a post-test measuring conceptual understanding, problem-solving, and transfer. Semi-structured interviews were also conducted to capture learner experiences.
Quantitative analysis included independent group comparisons using descriptive statistics, and inferential tests. Qualitative data were analyzed with inductive and deductive thematic analysis.
This study provides preliminary evidence that multi-representational instruction may improve problem-solving performance in Decision Tree learning. However, due to the small sample size and lack of statistical significance on two of three outcomes, these findings should be interpreted cautiously and require replication with larger samples.
The study contributes a structured evaluation framework for multi-representational ML education and provides evidence supporting the potential benefits of interactive instructional design in teaching Decision Trees.
Teaching Machines to Critique Computer Science Theses
A Human-in-the-Loop Framework for Discipline-Aware, Span-Anchored LLM Feedback in Thesis Supervision
To produce reviewable draft feedback on long computer-science thesis drafts under supervisor control, we present a controlled three-way comparison of LLM generation strategies on open-weight backbones, paired with a supervisor-facing PDF review interface. We compare base and LoRA-fine-tuned Llama 3.3 70B and Qwen 3.5 27B across whole-document, section-aware two-stage, and agentic review strategies, evaluated through an LLM-as-judge benchmark over all twelve configurations, a blind human rating study, and an interface usability study. Whole-document generation obtains the highest judge-macro score. Fine-tuning helps in only one of six matched base-versus-fine-tuned comparisons, indicating that supervised adaptation on a small span-anchored corpus shifts response distribution without expanding critique ability. In the blind study, the deployed section-aware generation is rated significantly higher than held-out supervisor feedback on change clarity (p = .027) and specificity (p = .018), while correctness shows no significant difference (p = .516). On the comments scored by both, three independent LLM judges rate the system above the human raters on every dimension (mean bias 0.41 points), most strongly on supervisor suitability. The PDF review interface reaches a mean System Usability Scale (SUS) score of 77.5. The results support using open-weight LLMs to produce reviewable draft feedback on long technical theses under supervisor inspection, editing, and export control.
...
To produce reviewable draft feedback on long computer-science thesis drafts under supervisor control, we present a controlled three-way comparison of LLM generation strategies on open-weight backbones, paired with a supervisor-facing PDF review interface. We compare base and LoRA-fine-tuned Llama 3.3 70B and Qwen 3.5 27B across whole-document, section-aware two-stage, and agentic review strategies, evaluated through an LLM-as-judge benchmark over all twelve configurations, a blind human rating study, and an interface usability study. Whole-document generation obtains the highest judge-macro score. Fine-tuning helps in only one of six matched base-versus-fine-tuned comparisons, indicating that supervised adaptation on a small span-anchored corpus shifts response distribution without expanding critique ability. In the blind study, the deployed section-aware generation is rated significantly higher than held-out supervisor feedback on change clarity (p = .027) and specificity (p = .018), while correctness shows no significant difference (p = .516). On the comments scored by both, three independent LLM judges rate the system above the human raters on every dimension (mean bias 0.41 points), most strongly on supervisor suitability. The PDF review interface reaches a mean System Usability Scale (SUS) score of 77.5. The results support using open-weight LLMs to produce reviewable draft feedback on long technical theses under supervisor inspection, editing, and export control.
Conceptual Bridges in Machine Learning
Exploring the Effect of Analogies on Multilayer Perceptron Understanding
Machine Learning education faces significant challenges due to the abstract and mathematically-complex nature of fundamental models, such as Multilayer Perceptrons (MLPs). This paper investigates the effectiveness of conceptual metaphors and analogies as pedagogical tools to improve novice learner's understanding of key MLP concepts. Using large language models, we generated a set of analogies for core MLP topics. These analogies were then evaluated by experts to assess their quality, followed by a user study with novice learners employing a between-subject A/B test comparing analogy-based explanations to formal definitions. Although the study found no statistically significant improvement in knowledge gain or engagement that could be attributed to analogy-based explanations, trends suggest potential benefits in learner confidence and motivation. The research contributes a curated set of expert-evaluated analogies for ML education and discusses methodology limitations and directions for future work. This study highlights both the promise and complexity of integrating analogy-based teaching approaches into ML education.
...
Machine Learning education faces significant challenges due to the abstract and mathematically-complex nature of fundamental models, such as Multilayer Perceptrons (MLPs). This paper investigates the effectiveness of conceptual metaphors and analogies as pedagogical tools to improve novice learner's understanding of key MLP concepts. Using large language models, we generated a set of analogies for core MLP topics. These analogies were then evaluated by experts to assess their quality, followed by a user study with novice learners employing a between-subject A/B test comparing analogy-based explanations to formal definitions. Although the study found no statistically significant improvement in knowledge gain or engagement that could be attributed to analogy-based explanations, trends suggest potential benefits in learner confidence and motivation. The research contributes a curated set of expert-evaluated analogies for ML education and discusses methodology limitations and directions for future work. This study highlights both the promise and complexity of integrating analogy-based teaching approaches into ML education.
Domain Specificity in Supervised Machine Learning Analogies
A Comparative Study of General Domain vs. Gaming Domain Analogies
This research paper looks into the influence of domain specificity on the understanding and motivation of first-year computer science students learning different concepts in supervised machine learning. Two types of domains were chosen for the analogies, the general domain and the gaming domain, the latter being the more specific one. These were evaluated in two phases. First, experts rated the analogies based on different metrics. Then, a user study was carried out using A/B testing to measure knowledge gain and motivation when exposed to the analogies. Results from the user evaluation show no statistically significant differences in terms of understanding for domain-specific or general analogies. Motivation, similarly show little difference when comparing both domains. The findings suggest that if analogies are helpful when it comes to understanding a topic, as long as the learner knows the domain, they do not play a big role.
...
This research paper looks into the influence of domain specificity on the understanding and motivation of first-year computer science students learning different concepts in supervised machine learning. Two types of domains were chosen for the analogies, the general domain and the gaming domain, the latter being the more specific one. These were evaluated in two phases. First, experts rated the analogies based on different metrics. Then, a user study was carried out using A/B testing to measure knowledge gain and motivation when exposed to the analogies. Results from the user evaluation show no statistically significant differences in terms of understanding for domain-specific or general analogies. Motivation, similarly show little difference when comparing both domains. The findings suggest that if analogies are helpful when it comes to understanding a topic, as long as the learner knows the domain, they do not play a big role.
How to Teach Unsupervised Machine Learning with Analogies
A Study on the Effectiveness of Analogies in Teaching Unsupervised Machine Learning
Unsupervised machine learning is a complex and abstract topic, posing challenges for student comprehension. Considering the considerable growth of relevance the topic of machine learning has seen in the past years, teaching it effectively has become ever-so important. Analogy-based teaching approaches offer a potential solution by mapping unfamiliar machine learning concepts to familiar real-world ideas. This paper investigates how analogies can improve the understanding of unsupervised learning, a rather relevant field within machine learning. Contributions include a collection of analogies for teaching unsupervised ML, an expert-based evaluation of these analogies’ quality, and a student-centred assessment of analogy-based teaching.
The findings from the expert evaluation show a consensus on the effectiveness of several analogies and highlight which analogies might be less effective. The findings from the student assessment suggest that the analogical explanations are more effective than 'generic' explanations and suggest that students have a higher satisfaction while learning through analogies.
We conclude that well-crafted analogies can enhance student understanding in unsupervised machine learning. The study’s insights can guide educators in integrating analogies to make unsupervised learning more accessible. ...
The findings from the expert evaluation show a consensus on the effectiveness of several analogies and highlight which analogies might be less effective. The findings from the student assessment suggest that the analogical explanations are more effective than 'generic' explanations and suggest that students have a higher satisfaction while learning through analogies.
We conclude that well-crafted analogies can enhance student understanding in unsupervised machine learning. The study’s insights can guide educators in integrating analogies to make unsupervised learning more accessible. ...
Unsupervised machine learning is a complex and abstract topic, posing challenges for student comprehension. Considering the considerable growth of relevance the topic of machine learning has seen in the past years, teaching it effectively has become ever-so important. Analogy-based teaching approaches offer a potential solution by mapping unfamiliar machine learning concepts to familiar real-world ideas. This paper investigates how analogies can improve the understanding of unsupervised learning, a rather relevant field within machine learning. Contributions include a collection of analogies for teaching unsupervised ML, an expert-based evaluation of these analogies’ quality, and a student-centred assessment of analogy-based teaching.
The findings from the expert evaluation show a consensus on the effectiveness of several analogies and highlight which analogies might be less effective. The findings from the student assessment suggest that the analogical explanations are more effective than 'generic' explanations and suggest that students have a higher satisfaction while learning through analogies.
We conclude that well-crafted analogies can enhance student understanding in unsupervised machine learning. The study’s insights can guide educators in integrating analogies to make unsupervised learning more accessible.
The findings from the expert evaluation show a consensus on the effectiveness of several analogies and highlight which analogies might be less effective. The findings from the student assessment suggest that the analogical explanations are more effective than 'generic' explanations and suggest that students have a higher satisfaction while learning through analogies.
We conclude that well-crafted analogies can enhance student understanding in unsupervised machine learning. The study’s insights can guide educators in integrating analogies to make unsupervised learning more accessible.
Teaching Gradient Descent Through Analogies, Step by Step
Evaluating and using analogies to teach concepts in Machine Learning to Computer Science students
Machine Learning is becoming a standard part of Computer Science curriculums at universities. This paper aims to contribute to the education of Machine Learning in Computer Science, specifically through teaching concepts related to Gradient Descent (GD) through analogies. First, concepts related to Gradient Descent were collected through the use of academic textbooks, and analogies were created based on the definitions found. These analogies were then evaluated by experts, scoring the analogies on Target Concept Coverage, Mapping Strength, and Metaphoricity. The analogies that scored highest on a mean average were then used in an A/B survey distributed amongst Computer Science students that had not followed any Machine Learning course. One group was given the concept definitions, the other both the definitions and the analogies. The learning proficiency was measured, and no statistically significant result was found. In the end, this research explores the possibilities of creating analogies to explain machine learning concepts, and provides a modular framework for evaluating quality and measuring effectiveness of analogies.
...
Machine Learning is becoming a standard part of Computer Science curriculums at universities. This paper aims to contribute to the education of Machine Learning in Computer Science, specifically through teaching concepts related to Gradient Descent (GD) through analogies. First, concepts related to Gradient Descent were collected through the use of academic textbooks, and analogies were created based on the definitions found. These analogies were then evaluated by experts, scoring the analogies on Target Concept Coverage, Mapping Strength, and Metaphoricity. The analogies that scored highest on a mean average were then used in an A/B survey distributed amongst Computer Science students that had not followed any Machine Learning course. One group was given the concept definitions, the other both the definitions and the analogies. The learning proficiency was measured, and no statistically significant result was found. In the end, this research explores the possibilities of creating analogies to explain machine learning concepts, and provides a modular framework for evaluating quality and measuring effectiveness of analogies.
This study examines the effect of analogies on conceptual understanding of machine learning (ML) loss functions, and the motivation to learn in first-year bachelor computer science students. For a set of 10 ML loss functions, analogies were generated and evaluated by 15 experts. 3 of these analogies were subsequently tested with 22 students. The results show no conclusive evidence for improvement in understanding and motivation to learn. The study outlines a general strategy for evaluation of analogies on student understanding and motivation. The study further provides 10 expert-rated analogies, 3 of which have been tested with students.
...
This study examines the effect of analogies on conceptual understanding of machine learning (ML) loss functions, and the motivation to learn in first-year bachelor computer science students. For a set of 10 ML loss functions, analogies were generated and evaluated by 15 experts. 3 of these analogies were subsequently tested with 22 students. The results show no conclusive evidence for improvement in understanding and motivation to learn. The study outlines a general strategy for evaluation of analogies on student understanding and motivation. The study further provides 10 expert-rated analogies, 3 of which have been tested with students.
Building Better Programmers: An AI System for Guided Program Decomposition
Analysing how guided program decomposition affects cognitive processes in computer science students
Generative AI has opened up new possibilities in computer science education. Large language models have made it possible for learners to get instantaneous and customised feedback on different programming concepts, as well as the ability to use natural language to implement these concepts. One such concept is program decomposition, an essential skill in software engineering. This work presents a novel method for teaching program decomposition, using a three-stage guided AI decomposition system. We analyse how this method affects a learner's program decomposition cognitive processes via a concurrent think-aloud protocol where a student decomposes three simple programming tasks. Furthermore, we measure how using the system changes a student's confidence in their decomposition skills. We find that participants do not display any significant change in confidence levels after using the system. We observe that the students display a significant improvement in performance during the course of the study. The participants also display a significant decrease in metacognitive confusion and a clear emergence of reflection based on previous errors. We conclude that the proposed method and the implemented system lead to a level of internalisation of decomposition skills in the students. We recommend that a study of change in decomposition skills is conducted over a longer time period to observe the full effects of the method.
...
Generative AI has opened up new possibilities in computer science education. Large language models have made it possible for learners to get instantaneous and customised feedback on different programming concepts, as well as the ability to use natural language to implement these concepts. One such concept is program decomposition, an essential skill in software engineering. This work presents a novel method for teaching program decomposition, using a three-stage guided AI decomposition system. We analyse how this method affects a learner's program decomposition cognitive processes via a concurrent think-aloud protocol where a student decomposes three simple programming tasks. Furthermore, we measure how using the system changes a student's confidence in their decomposition skills. We find that participants do not display any significant change in confidence levels after using the system. We observe that the students display a significant improvement in performance during the course of the study. The participants also display a significant decrease in metacognitive confusion and a clear emergence of reflection based on previous errors. We conclude that the proposed method and the implemented system lead to a level of internalisation of decomposition skills in the students. We recommend that a study of change in decomposition skills is conducted over a longer time period to observe the full effects of the method.
Comparing the hint quality of a Small Language Model and a Large Language Model in automatic hint generation
Replacing the LLM inside the JetBrains Academy AI hint generation system with a RAG-augmented SLM
The rapid advancement of Large Language Models (LLMs) in recent years is not without concerns, such as a lack of privacy, environmental impact, and financial concerns. It might therefore be beneficial to use Small Language Models (SLMs) instead, which are more accessible to be run by individuals or organisations, thus resulting in more control over the model. This research investigates whether we can replace an LLM with an SLM inside an AI hint-generation system, and achieve comparable hint quality, by conducting an expert study to validate generated hints based on a set of criteria and by conducting a student experiment, investigating student satisfaction and trust in the system. The expert results show that the hints generated by the SLM-powered system are slightly less personalised to the situation, are noticeably more misleading and more often suggest the wrong approach. The student experiment shows similar results for these criteria, and shows a slight decrease in the overall perceived helpfulness of the hints, trust in the system and willingness to continue using the system. The most prevalent complaint for the SLM-powered system was its inconsistency in the hint quality, as it generated good and useful hints in some contexts, but also suggested wrong and unusable hints too often. Thus, while replacing the LLM with an SLM has potential, as it is capable of generating useful hints, current SLMs are still too inconsistent.
...
The rapid advancement of Large Language Models (LLMs) in recent years is not without concerns, such as a lack of privacy, environmental impact, and financial concerns. It might therefore be beneficial to use Small Language Models (SLMs) instead, which are more accessible to be run by individuals or organisations, thus resulting in more control over the model. This research investigates whether we can replace an LLM with an SLM inside an AI hint-generation system, and achieve comparable hint quality, by conducting an expert study to validate generated hints based on a set of criteria and by conducting a student experiment, investigating student satisfaction and trust in the system. The expert results show that the hints generated by the SLM-powered system are slightly less personalised to the situation, are noticeably more misleading and more often suggest the wrong approach. The student experiment shows similar results for these criteria, and shows a slight decrease in the overall perceived helpfulness of the hints, trust in the system and willingness to continue using the system. The most prevalent complaint for the SLM-powered system was its inconsistency in the hint quality, as it generated good and useful hints in some contexts, but also suggested wrong and unusable hints too often. Thus, while replacing the LLM with an SLM has potential, as it is capable of generating useful hints, current SLMs are still too inconsistent.
Learning Machine Learning: A Comparative Study of Industrial Design and Computer Science and Engineering Students
Exploring the Role of Mathematics Backgrounds in Foundational ML Education
Machine learning (ML) has become a critical skill across various disciplines, yet teaching it to students outside Computer Science and Engineering (CS) remains challenging due to differing academic backgrounds. This study investigates the differences in learning outcomes between Industrial Design (ID) and CS students when introduced to foundational ML topics, focusing on the influence of prior mathematical knowledge.
Through initial surveys on mathematical proficiency, structured ML tutorials, and final assessments on learning outcomes, the research examines correlations between mathematical proficiency and ML performance, faculty-specific challenges, and qualitative feedback on learning experiences. Results reveal that prior mathematics knowledge significantly impacts performance on mathematics-intensive topics such as Bayes' Rule, while its influence is minimal on less math-relevant topics like ML pipelines. Furthermore, ID students emphasized creative and interactive teaching methods, contrasting with the programming-oriented preferences of CS students.
These findings highlight the need for interdisciplinary instructional strategies that cater to diverse learner strengths. By uncovering faculty-specific patterns in ML learning, this study contributes to the design of more inclusive and effective educational practices, fostering a broader understanding and application of ML across disciplines. ...
Through initial surveys on mathematical proficiency, structured ML tutorials, and final assessments on learning outcomes, the research examines correlations between mathematical proficiency and ML performance, faculty-specific challenges, and qualitative feedback on learning experiences. Results reveal that prior mathematics knowledge significantly impacts performance on mathematics-intensive topics such as Bayes' Rule, while its influence is minimal on less math-relevant topics like ML pipelines. Furthermore, ID students emphasized creative and interactive teaching methods, contrasting with the programming-oriented preferences of CS students.
These findings highlight the need for interdisciplinary instructional strategies that cater to diverse learner strengths. By uncovering faculty-specific patterns in ML learning, this study contributes to the design of more inclusive and effective educational practices, fostering a broader understanding and application of ML across disciplines. ...
Machine learning (ML) has become a critical skill across various disciplines, yet teaching it to students outside Computer Science and Engineering (CS) remains challenging due to differing academic backgrounds. This study investigates the differences in learning outcomes between Industrial Design (ID) and CS students when introduced to foundational ML topics, focusing on the influence of prior mathematical knowledge.
Through initial surveys on mathematical proficiency, structured ML tutorials, and final assessments on learning outcomes, the research examines correlations between mathematical proficiency and ML performance, faculty-specific challenges, and qualitative feedback on learning experiences. Results reveal that prior mathematics knowledge significantly impacts performance on mathematics-intensive topics such as Bayes' Rule, while its influence is minimal on less math-relevant topics like ML pipelines. Furthermore, ID students emphasized creative and interactive teaching methods, contrasting with the programming-oriented preferences of CS students.
These findings highlight the need for interdisciplinary instructional strategies that cater to diverse learner strengths. By uncovering faculty-specific patterns in ML learning, this study contributes to the design of more inclusive and effective educational practices, fostering a broader understanding and application of ML across disciplines.
Through initial surveys on mathematical proficiency, structured ML tutorials, and final assessments on learning outcomes, the research examines correlations between mathematical proficiency and ML performance, faculty-specific challenges, and qualitative feedback on learning experiences. Results reveal that prior mathematics knowledge significantly impacts performance on mathematics-intensive topics such as Bayes' Rule, while its influence is minimal on less math-relevant topics like ML pipelines. Furthermore, ID students emphasized creative and interactive teaching methods, contrasting with the programming-oriented preferences of CS students.
These findings highlight the need for interdisciplinary instructional strategies that cater to diverse learner strengths. By uncovering faculty-specific patterns in ML learning, this study contributes to the design of more inclusive and effective educational practices, fostering a broader understanding and application of ML across disciplines.
Advantages of Prior Mathematical Knowledge for Studying Machine Learning
Differences in Knowledge Gain between Computer Science and Physics Students
With the growing need for machine learning knowledge for many different expertises and positions, comes a growing need for machine learning education for non-computer scientists. Teaching machine learning concepts to non-majors comes with the added challenge of dealing with different levels of prior mathematical knowledge. Existing research is inconclusive on the correlation between this prior knowledge and topic-specific machine learning knowledge gain. This paper evaluated this via an experiment conducted on Computer Science and Physics students without prior machine learning education. We find that there is no clear correlation between general math knowledge and knowledge gain. There is however a clear correlation of proficiency in probability and statistics, and algorithm heavy machine learning topics. The experiment also concluded that most students struggled most with these math-heavy topics, as well as understanding abstract systems such as perceptrons.
...
With the growing need for machine learning knowledge for many different expertises and positions, comes a growing need for machine learning education for non-computer scientists. Teaching machine learning concepts to non-majors comes with the added challenge of dealing with different levels of prior mathematical knowledge. Existing research is inconclusive on the correlation between this prior knowledge and topic-specific machine learning knowledge gain. This paper evaluated this via an experiment conducted on Computer Science and Physics students without prior machine learning education. We find that there is no clear correlation between general math knowledge and knowledge gain. There is however a clear correlation of proficiency in probability and statistics, and algorithm heavy machine learning topics. The experiment also concluded that most students struggled most with these math-heavy topics, as well as understanding abstract systems such as perceptrons.
Learning Machine Learning
A Comparative Study of Aerospace Engineering and Computer Science Students
Machine learning (ML) is increasingly integrated across diverse academic disciplines, necessitating effective teaching strategies tailored to varied student backgrounds. This study investigates the influence of prior mathematical knowledge on the learning outcomes of ML topics among Computer Science (CS) and Aerospace Engineering (AE) students. Employing a mixed-methods approach, the research involved initial mathematical assessments, interactive tutorials on key ML topics (Bayes Rule, Perceptrons, ML Pipelines), and subsequent evaluations of ML comprehension.
The results reveal significant differences in performance between the two groups. CS students, with their integrated programming and mathematical preparation, consistently outperformed AE students, who demonstrated variability despite their strong quantitative foundations. Probability and linear algebra emerged as key contributors to ML learning, showing stronger correlations with outcomes than calculus. Qualitative analysis highlighted the need for tailored instructional approaches: AE students preferred application-driven and interactive learning, while CS students valued structured and technically detailed resources.
These findings underscore the importance of interdisciplinary teaching strategies that bridge gaps in programming and mathematical competencies. The study’s insights have implications for designing inclusive ML curricula, emphasizing real-world applications, adaptive learning technologies, and frameworks to support diverse learner needs. Future research should explore broader ML topics, larger participant groups, and long-term skill retention to further enhance ML education across disciplines. ...
The results reveal significant differences in performance between the two groups. CS students, with their integrated programming and mathematical preparation, consistently outperformed AE students, who demonstrated variability despite their strong quantitative foundations. Probability and linear algebra emerged as key contributors to ML learning, showing stronger correlations with outcomes than calculus. Qualitative analysis highlighted the need for tailored instructional approaches: AE students preferred application-driven and interactive learning, while CS students valued structured and technically detailed resources.
These findings underscore the importance of interdisciplinary teaching strategies that bridge gaps in programming and mathematical competencies. The study’s insights have implications for designing inclusive ML curricula, emphasizing real-world applications, adaptive learning technologies, and frameworks to support diverse learner needs. Future research should explore broader ML topics, larger participant groups, and long-term skill retention to further enhance ML education across disciplines. ...
Machine learning (ML) is increasingly integrated across diverse academic disciplines, necessitating effective teaching strategies tailored to varied student backgrounds. This study investigates the influence of prior mathematical knowledge on the learning outcomes of ML topics among Computer Science (CS) and Aerospace Engineering (AE) students. Employing a mixed-methods approach, the research involved initial mathematical assessments, interactive tutorials on key ML topics (Bayes Rule, Perceptrons, ML Pipelines), and subsequent evaluations of ML comprehension.
The results reveal significant differences in performance between the two groups. CS students, with their integrated programming and mathematical preparation, consistently outperformed AE students, who demonstrated variability despite their strong quantitative foundations. Probability and linear algebra emerged as key contributors to ML learning, showing stronger correlations with outcomes than calculus. Qualitative analysis highlighted the need for tailored instructional approaches: AE students preferred application-driven and interactive learning, while CS students valued structured and technically detailed resources.
These findings underscore the importance of interdisciplinary teaching strategies that bridge gaps in programming and mathematical competencies. The study’s insights have implications for designing inclusive ML curricula, emphasizing real-world applications, adaptive learning technologies, and frameworks to support diverse learner needs. Future research should explore broader ML topics, larger participant groups, and long-term skill retention to further enhance ML education across disciplines.
The results reveal significant differences in performance between the two groups. CS students, with their integrated programming and mathematical preparation, consistently outperformed AE students, who demonstrated variability despite their strong quantitative foundations. Probability and linear algebra emerged as key contributors to ML learning, showing stronger correlations with outcomes than calculus. Qualitative analysis highlighted the need for tailored instructional approaches: AE students preferred application-driven and interactive learning, while CS students valued structured and technically detailed resources.
These findings underscore the importance of interdisciplinary teaching strategies that bridge gaps in programming and mathematical competencies. The study’s insights have implications for designing inclusive ML curricula, emphasizing real-world applications, adaptive learning technologies, and frameworks to support diverse learner needs. Future research should explore broader ML topics, larger participant groups, and long-term skill retention to further enhance ML education across disciplines.