TN
T.T. Naber
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 this thesis, we study the impact of sparsity on both the performance and interpretability of Graph Attention Networks. Additionally, we introduce two novel methods that yield Sparse & Interpretable Graph Attention Networks.
In Chapter 1, we introduce the reader to the concept of Graph Attention Networks (GATs), sparse attention, and interpretability. Subsequently, we provide our research motivation and formulate the research question.
In Chapter 2, the necessary background information is presented at a fundamental level. First, the reader is formally introduced to Graph Neural Networks (GNNs) and GATs. Continuing, various approaches to sparse attention are discussed in detail. Finally, an overview of the current approaches to explainability in GNNs is provided, along with a focus on the category that the methods in this research can be assigned to.
Chapter 3 consists of the paper written on this topic. As this report acts as an envelope of the paper, the introduction and background of the paper overlap with the corresponding chapters in this report, although the paper is written more formally and concisely. Most importantly, this chapter contains an in-depth explanation of the methods developed in this study and their evaluation. Furthermore, rigorous evaluations are performed and presented in the form of Pareto curves, providing insights into the performance-interpretability trade-off for all datasets. Finally, an appendix is provided containing details regarding the evaluation and some additional results.
The paper has to function as a stand-alone research and is therefore considered the core of this report. However, a paper has its limitations due to its concise nature. Thus, Chapter 4 contains additional evaluations performed to gain insight into the behaviour of sparsity within the proposed methods and the effect of changing the sparsity parameter after training. Furthermore, we presented a failed concept due to its relevance for future work. Additional related work is presented in Chapter 5, where we discuss the idea of attention as an explanation and present other self-interpretable methods within the field.
This research is concluded by providing an answer to the research question in Chapter 6, along with suggestions for future work. ...
In Chapter 1, we introduce the reader to the concept of Graph Attention Networks (GATs), sparse attention, and interpretability. Subsequently, we provide our research motivation and formulate the research question.
In Chapter 2, the necessary background information is presented at a fundamental level. First, the reader is formally introduced to Graph Neural Networks (GNNs) and GATs. Continuing, various approaches to sparse attention are discussed in detail. Finally, an overview of the current approaches to explainability in GNNs is provided, along with a focus on the category that the methods in this research can be assigned to.
Chapter 3 consists of the paper written on this topic. As this report acts as an envelope of the paper, the introduction and background of the paper overlap with the corresponding chapters in this report, although the paper is written more formally and concisely. Most importantly, this chapter contains an in-depth explanation of the methods developed in this study and their evaluation. Furthermore, rigorous evaluations are performed and presented in the form of Pareto curves, providing insights into the performance-interpretability trade-off for all datasets. Finally, an appendix is provided containing details regarding the evaluation and some additional results.
The paper has to function as a stand-alone research and is therefore considered the core of this report. However, a paper has its limitations due to its concise nature. Thus, Chapter 4 contains additional evaluations performed to gain insight into the behaviour of sparsity within the proposed methods and the effect of changing the sparsity parameter after training. Furthermore, we presented a failed concept due to its relevance for future work. Additional related work is presented in Chapter 5, where we discuss the idea of attention as an explanation and present other self-interpretable methods within the field.
This research is concluded by providing an answer to the research question in Chapter 6, along with suggestions for future work. ...
In this thesis, we study the impact of sparsity on both the performance and interpretability of Graph Attention Networks. Additionally, we introduce two novel methods that yield Sparse & Interpretable Graph Attention Networks.
In Chapter 1, we introduce the reader to the concept of Graph Attention Networks (GATs), sparse attention, and interpretability. Subsequently, we provide our research motivation and formulate the research question.
In Chapter 2, the necessary background information is presented at a fundamental level. First, the reader is formally introduced to Graph Neural Networks (GNNs) and GATs. Continuing, various approaches to sparse attention are discussed in detail. Finally, an overview of the current approaches to explainability in GNNs is provided, along with a focus on the category that the methods in this research can be assigned to.
Chapter 3 consists of the paper written on this topic. As this report acts as an envelope of the paper, the introduction and background of the paper overlap with the corresponding chapters in this report, although the paper is written more formally and concisely. Most importantly, this chapter contains an in-depth explanation of the methods developed in this study and their evaluation. Furthermore, rigorous evaluations are performed and presented in the form of Pareto curves, providing insights into the performance-interpretability trade-off for all datasets. Finally, an appendix is provided containing details regarding the evaluation and some additional results.
The paper has to function as a stand-alone research and is therefore considered the core of this report. However, a paper has its limitations due to its concise nature. Thus, Chapter 4 contains additional evaluations performed to gain insight into the behaviour of sparsity within the proposed methods and the effect of changing the sparsity parameter after training. Furthermore, we presented a failed concept due to its relevance for future work. Additional related work is presented in Chapter 5, where we discuss the idea of attention as an explanation and present other self-interpretable methods within the field.
This research is concluded by providing an answer to the research question in Chapter 6, along with suggestions for future work.
In Chapter 1, we introduce the reader to the concept of Graph Attention Networks (GATs), sparse attention, and interpretability. Subsequently, we provide our research motivation and formulate the research question.
In Chapter 2, the necessary background information is presented at a fundamental level. First, the reader is formally introduced to Graph Neural Networks (GNNs) and GATs. Continuing, various approaches to sparse attention are discussed in detail. Finally, an overview of the current approaches to explainability in GNNs is provided, along with a focus on the category that the methods in this research can be assigned to.
Chapter 3 consists of the paper written on this topic. As this report acts as an envelope of the paper, the introduction and background of the paper overlap with the corresponding chapters in this report, although the paper is written more formally and concisely. Most importantly, this chapter contains an in-depth explanation of the methods developed in this study and their evaluation. Furthermore, rigorous evaluations are performed and presented in the form of Pareto curves, providing insights into the performance-interpretability trade-off for all datasets. Finally, an appendix is provided containing details regarding the evaluation and some additional results.
The paper has to function as a stand-alone research and is therefore considered the core of this report. However, a paper has its limitations due to its concise nature. Thus, Chapter 4 contains additional evaluations performed to gain insight into the behaviour of sparsity within the proposed methods and the effect of changing the sparsity parameter after training. Furthermore, we presented a failed concept due to its relevance for future work. Additional related work is presented in Chapter 5, where we discuss the idea of attention as an explanation and present other self-interpretable methods within the field.
This research is concluded by providing an answer to the research question in Chapter 6, along with suggestions for future work.
Bachelor thesis
(2020)
-
Frank Bredius, Titus Naber, Bailey Tjiong, Leroy Velzel, Kin-Fai Chan, A Katsifodimos, Otto Visser
Scenwise has developed the ScenarioDesigner to support road authorities in creating response plans. The response plans describe which actions the operator should take in order to manage different traffic situations such as accidents, large scale events This thesis elaborates on the process as well as the end-product created during the Bachelor End Project The BEP is a 10 week compulsory project to complete the Computer Science and Engineering program of the Delft University of Technology. It was conducted between 11 November 2019 and 11 February 2020. During the first two weeks the group focused on researching the problem, the possible solutions and traffic management in general. The subsequent 8 weeks were designated to the development of the product. The product was developed for ScenWise, a company that focuses on innovation within traffic management. In addition to evaluating the product, this thesis gives recommendations for a following group working on extending the product.
Finally, the group would like to thank the client ScenWise, specifically K.F. Chan for providing the project as well as teaching us about the domain knowledge of the client. Furthermore, the group wants to thank Dr. A. Katsifodimos for the coaching throughout the project as well as Ir. O.W. Visser for giving the opportunity to do the BEP in the second quarter of the academic year. ...
Finally, the group would like to thank the client ScenWise, specifically K.F. Chan for providing the project as well as teaching us about the domain knowledge of the client. Furthermore, the group wants to thank Dr. A. Katsifodimos for the coaching throughout the project as well as Ir. O.W. Visser for giving the opportunity to do the BEP in the second quarter of the academic year. ...
Scenwise has developed the ScenarioDesigner to support road authorities in creating response plans. The response plans describe which actions the operator should take in order to manage different traffic situations such as accidents, large scale events This thesis elaborates on the process as well as the end-product created during the Bachelor End Project The BEP is a 10 week compulsory project to complete the Computer Science and Engineering program of the Delft University of Technology. It was conducted between 11 November 2019 and 11 February 2020. During the first two weeks the group focused on researching the problem, the possible solutions and traffic management in general. The subsequent 8 weeks were designated to the development of the product. The product was developed for ScenWise, a company that focuses on innovation within traffic management. In addition to evaluating the product, this thesis gives recommendations for a following group working on extending the product.
Finally, the group would like to thank the client ScenWise, specifically K.F. Chan for providing the project as well as teaching us about the domain knowledge of the client. Furthermore, the group wants to thank Dr. A. Katsifodimos for the coaching throughout the project as well as Ir. O.W. Visser for giving the opportunity to do the BEP in the second quarter of the academic year.
Finally, the group would like to thank the client ScenWise, specifically K.F. Chan for providing the project as well as teaching us about the domain knowledge of the client. Furthermore, the group wants to thank Dr. A. Katsifodimos for the coaching throughout the project as well as Ir. O.W. Visser for giving the opportunity to do the BEP in the second quarter of the academic year.