Abstract
Learning from small data sets is difficult in the absence of specific domain knowledge. We present a regularized linear model called STEW that benefits from a generic and prevalent form of prior knowledge: feature directions. STEW shrinks weights toward each other, converging to an equal-weights solution in the limit of infinite regularization. We provide theoretical results on the equal-weights solution that explains how STEW can productively trade-off bias and variance. Across a wide range of learning problems, including Tetris, STEW outperformed existing linear models, including ridge regression, the Lasso, and the non-negative Lasso, when feature directions were known. The model proved to be robust to unreliable (or absent) feature directions, still outperforming alternative models under diverse conditions. Our results in Tetris were obtained by using a novel approach to learning in sequential decision environments based on multinomial logistic regression.
This project has received funding from the European Union’s Horizon 2020 research and innovation programme under the Marie Skłodowska-Curie grant agreement No 665992
This project has received funding from the European Union’s Horizon 2020 research and innovation programme under the Marie Skłodowska-Curie grant agreement No 665992
Original language | English |
---|---|
Title of host publication | Proceedings of Machine Learning Research |
Publisher | International Machine Learning Society (IMLS) |
Pages | 3953-3962 |
Number of pages | 10 |
Volume | 97 |
Publication status | Published - 15 Jun 2019 |
Event | Thirty-sixth International Conference on Machine Learning - Long Beach Convention Center, Long Beach, USA United States Duration: 9 Jun 2019 → 15 Jun 2019 Conference number: 36 https://icml.cc/ |
Publication series
Name | Proceedings of Machine Learning Research |
---|---|
Publisher | International Machine Learning Society (IMLS) |
Volume | 97 |
ISSN (Electronic) | 2640-3498 |
Conference
Conference | Thirty-sixth International Conference on Machine Learning |
---|---|
Abbreviated title | ICML |
Country/Territory | USA United States |
City | Long Beach |
Period | 9/06/19 → 15/06/19 |
Internet address |
Keywords
- Machine learning
- Reinforcement learning
- Regularization
- Equal weights
Fingerprint
Dive into the research topics of 'Regularization in Directable Environments with Application to Tetris'. Together they form a unique fingerprint.Profiles
-
Özgür Şimşek
- Department of Computer Science - Deputy Head of Department
- UKRI CDT in Accountable, Responsible and Transparent AI
- Centre for Mathematics and Algorithms for Data (MAD)
- Artificial Intelligence and Machine Learning - Head of Group
- EPSRC Centre for Doctoral Training in Statistical Applied Mathematics (SAMBa)
- Bath Institute for the Augmented Human
Person: Research & Teaching, Core staff