Deep neural networks are lazy : on the inductive bias of deep learning
This electronic version was submitted by the student author. The certified thesis is available in the Institute Archives and Special Collections.
| Main Author: | |
|---|---|
| Other Authors: | |
| Format: | Thesis |
| Language: | eng |
| Published: |
Massachusetts Institute of Technology
2019
|
| Subjects: | |
| Online Access: | https://hdl.handle.net/1721.1/121680 |
| _version_ | 1811077302241460224 |
|---|---|
| author | Mansour, Tarek,M. Eng.Massachusetts Institute of Technology. |
| author2 | Aleksander Madry. |
| author_facet | Aleksander Madry. Mansour, Tarek,M. Eng.Massachusetts Institute of Technology. |
| author_sort | Mansour, Tarek,M. Eng.Massachusetts Institute of Technology. |
| collection | MIT |
| description | This electronic version was submitted by the student author. The certified thesis is available in the Institute Archives and Special Collections. |
| first_indexed | 2024-09-23T10:40:48Z |
| format | Thesis |
| id | mit-1721.1/121680 |
| institution | Massachusetts Institute of Technology |
| language | eng |
| last_indexed | 2024-09-23T10:40:48Z |
| publishDate | 2019 |
| publisher | Massachusetts Institute of Technology |
| record_format | dspace |
| spelling | mit-1721.1/1216802019-08-07T03:04:52Z Deep neural networks are lazy : on the inductive bias of deep learning On the inductive bias of deep learning Mansour, Tarek,M. Eng.Massachusetts Institute of Technology. Aleksander Madry. Massachusetts Institute of Technology. Department of Electrical Engineering and Computer Science. Massachusetts Institute of Technology. Department of Electrical Engineering and Computer Science Electrical Engineering and Computer Science. This electronic version was submitted by the student author. The certified thesis is available in the Institute Archives and Special Collections. Thesis: M. Eng., Massachusetts Institute of Technology, Department of Electrical Engineering and Computer Science, 2019 Cataloged from student-submitted PDF version of thesis. Includes bibliographical references (pages 75-78). Deep learning models exhibit superior generalization performance despite being heavily overparametrized. Although widely observed in practice, there is currently very little theoretical backing for such a phenomena. In this thesis, we propose a step forward towards understanding generalization in deep learning. We present evidence that deep neural networks have an inherent inductive bias that makes them inclined to learn generalizable hypotheses and avoid memorization. In this respect, we propose results that suggest that the inductive bias stems from neural networks being lazy: they tend to learn simpler rules first. We also propose a definition of simplicity in deep learning based on the implicit priors ingrained in deep neural networks. by Tarek Mansour. M. Eng. M.Eng. Massachusetts Institute of Technology, Department of Electrical Engineering and Computer Science 2019-07-15T20:33:32Z 2019-07-15T20:33:32Z 2019 2019 Thesis https://hdl.handle.net/1721.1/121680 1102057114 eng MIT theses are protected by copyright. They may be viewed, downloaded, or printed from this source but further reproduction or distribution in any format is prohibited without written permission. http://dspace.mit.edu/handle/1721.1/7582 78 pages application/pdf Massachusetts Institute of Technology |
| spellingShingle | Electrical Engineering and Computer Science. Mansour, Tarek,M. Eng.Massachusetts Institute of Technology. Deep neural networks are lazy : on the inductive bias of deep learning |
| title | Deep neural networks are lazy : on the inductive bias of deep learning |
| title_full | Deep neural networks are lazy : on the inductive bias of deep learning |
| title_fullStr | Deep neural networks are lazy : on the inductive bias of deep learning |
| title_full_unstemmed | Deep neural networks are lazy : on the inductive bias of deep learning |
| title_short | Deep neural networks are lazy : on the inductive bias of deep learning |
| title_sort | deep neural networks are lazy on the inductive bias of deep learning |
| topic | Electrical Engineering and Computer Science. |
| url | https://hdl.handle.net/1721.1/121680 |
| work_keys_str_mv | AT mansourtarekmengmassachusettsinstituteoftechnology deepneuralnetworksarelazyontheinductivebiasofdeeplearning AT mansourtarekmengmassachusettsinstituteoftechnology ontheinductivebiasofdeeplearning |