SN
S.C. Noorthoek
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>
2 records found
1
In addition to delivering groceries at customers’ doorsteps, online supermarket Picnic goes the extra mile by aiming to improve customer satisfaction. For instance, by providing cooking inspiration to customers through a recently launched recipe page in the app. This feature presents new recipes weekly and allows customers to easily add the ingredients to their shopping basket. It has raised interest in finding out what dishes customers are cooking as it could be helpful in choosing recipes for the page, predicting which articles are forgotten before checkout, and building a recipe recommender system. Hence, this work proposes two models to detect dish types in Picnic deliveries. The problem is scoped to detect main meals from a specified list of dish types in deliveries which were ordered in the Netherlands. The first model, named the Frequent Itemset Model, applies unsupervised learning techniques. First the articles in the deliveries are pre-processed by removing certain articles, choosing the representation of articles, and cleaning the text. The itemsets which represent core ingredients are obtained by applying techniques such as frequent itemset mining, association rule mining, and hierarchical clustering. In the final step itemsets are matched to dish types with the use of programmatic labelling and fuzzy string matching. Newly available labelled data enables the creation of a second model, referred to as the Supervised Learning Model, which applies supervised learning techniques. Features are selected and extracted. Multiple machine learning models, some in combination with binary relevance, are compared. The models are evaluated on two datasets: a large, weakly labelled dataset obtained through the recipe page, and a very small, manually labelled dataset with deliveries from a single customer. It is challenging to evaluate the performance of the models, since no large, truly labelled dataset is available. The results do indicate that the Frequent Itemset Model is able to detect common dish types, and that the Supervised Learning Model is able to detect dish types which are similar to the Picnic recipes it has trained on. Multiple suggestions are made for future work, such as obtaining a larger variety of labelled data and redefining the class labels. The contribution of this work is the formulation of the problem, two proposed solutions, insights into the challenges, and suggestions for future work.
...
In addition to delivering groceries at customers’ doorsteps, online supermarket Picnic goes the extra mile by aiming to improve customer satisfaction. For instance, by providing cooking inspiration to customers through a recently launched recipe page in the app. This feature presents new recipes weekly and allows customers to easily add the ingredients to their shopping basket. It has raised interest in finding out what dishes customers are cooking as it could be helpful in choosing recipes for the page, predicting which articles are forgotten before checkout, and building a recipe recommender system. Hence, this work proposes two models to detect dish types in Picnic deliveries. The problem is scoped to detect main meals from a specified list of dish types in deliveries which were ordered in the Netherlands. The first model, named the Frequent Itemset Model, applies unsupervised learning techniques. First the articles in the deliveries are pre-processed by removing certain articles, choosing the representation of articles, and cleaning the text. The itemsets which represent core ingredients are obtained by applying techniques such as frequent itemset mining, association rule mining, and hierarchical clustering. In the final step itemsets are matched to dish types with the use of programmatic labelling and fuzzy string matching. Newly available labelled data enables the creation of a second model, referred to as the Supervised Learning Model, which applies supervised learning techniques. Features are selected and extracted. Multiple machine learning models, some in combination with binary relevance, are compared. The models are evaluated on two datasets: a large, weakly labelled dataset obtained through the recipe page, and a very small, manually labelled dataset with deliveries from a single customer. It is challenging to evaluate the performance of the models, since no large, truly labelled dataset is available. The results do indicate that the Frequent Itemset Model is able to detect common dish types, and that the Supervised Learning Model is able to detect dish types which are similar to the Picnic recipes it has trained on. Multiple suggestions are made for future work, such as obtaining a larger variety of labelled data and redefining the class labels. The contribution of this work is the formulation of the problem, two proposed solutions, insights into the challenges, and suggestions for future work.
Bachelor thesis
(2019)
-
Maxim Liefaard, Raoul Bruens, Dana van Hassel, Sterre Noorthoek, Thomas Abeel, Maneesh Verma, Chris Verhoeven, Otto Visser, Huijuan Wang
With the steady increase in space missions, enabled through technological advances and increase of commercialisation within the space flight industry, both more and increasingly complex missions can be designed for space. To this end, the Lunar Zebro project competes within this field through its small lunar rover design, drastically decreasing deployment costs and risk of the mission. The road map of Lunar Zebro aims to have a multitude of rovers deployed on the Moon, being able to complete several tasks like exploring, observing, and mapping. Since this concept of rover cooperation adds a novel level of complexity to the mission, a feasibility study is required to look into the difficulties of navigating the Moon with a larger group of rovers. LunarSim is the software package developed during this project. LunarSim aims to facilitate a simulation environment in which Lunar Zebro rovers and space mission designs can be tested and validated. To legitimise the workings of the simulation, a few scenarios have been developed to test the core functionalities of the software product. These scenarios are based on phases in a practical mission plan that consists out of navigating to and observing a crater location. The scenarios is evaluated through examination of a set of defined fitness criteria. In this report, the reader will find documentation on the development process of LunarSim: the simulation in Unity, the ROS back-end, and the bridge between these two systems. Additionally, the report elaborates how the developed software was used to aid in the feasibility study of LUFAR. First, initial research and requirements are formulated to define the scope of the simulation, after which the software architecture is introduced. Then, the systems implemented for the simulation are explained. Subsequently, the implemented rover behaviour algorithm that was used for testing is explained, with additional resources on how to develop a new custom rover behaviour. After this, an evaluation is given of the simulation based on the initial requirements and research with future research and concluding remarks. At the end of the report, the technical specifications in terms of software architecture, simulation environment, and rover behaviour are defined to give an in-depth view of LunarSim.
...
With the steady increase in space missions, enabled through technological advances and increase of commercialisation within the space flight industry, both more and increasingly complex missions can be designed for space. To this end, the Lunar Zebro project competes within this field through its small lunar rover design, drastically decreasing deployment costs and risk of the mission. The road map of Lunar Zebro aims to have a multitude of rovers deployed on the Moon, being able to complete several tasks like exploring, observing, and mapping. Since this concept of rover cooperation adds a novel level of complexity to the mission, a feasibility study is required to look into the difficulties of navigating the Moon with a larger group of rovers. LunarSim is the software package developed during this project. LunarSim aims to facilitate a simulation environment in which Lunar Zebro rovers and space mission designs can be tested and validated. To legitimise the workings of the simulation, a few scenarios have been developed to test the core functionalities of the software product. These scenarios are based on phases in a practical mission plan that consists out of navigating to and observing a crater location. The scenarios is evaluated through examination of a set of defined fitness criteria. In this report, the reader will find documentation on the development process of LunarSim: the simulation in Unity, the ROS back-end, and the bridge between these two systems. Additionally, the report elaborates how the developed software was used to aid in the feasibility study of LUFAR. First, initial research and requirements are formulated to define the scope of the simulation, after which the software architecture is introduced. Then, the systems implemented for the simulation are explained. Subsequently, the implemented rover behaviour algorithm that was used for testing is explained, with additional resources on how to develop a new custom rover behaviour. After this, an evaluation is given of the simulation based on the initial requirements and research with future research and concluding remarks. At the end of the report, the technical specifications in terms of software architecture, simulation environment, and rover behaviour are defined to give an in-depth view of LunarSim.