Description
| Title | Overcoming the Cold Start Problem for Personalized Reinforcement Learning |
| Abstract | A problem that all data-driven approaches for personalization suffer from is the cold start problem: in the beginning, very few experiences with the user are available, making it nearly impossible to provide a good level of personalization while this is precisely the period where the user should become engaged. In this talk, I will explain a new approach that improves the speed and the performance of reinforcement learning in the context of personalization. |
Other presentations by Amin Tabatabaei
| Date | Title |
|---|---|
| 02 February 2015 | Mathematical Analysis of Smart Daily Energy Management for Heat Pumps |
| 12 October 2015 | |
| 09 May 2016 | |
| 30 January 2017 | A Data Analysis Approach for Diagnosing Malfunctioning in Domestic Space Heating |
| 24 September 2018 | Overcoming the Cold Start Problem for Personalized Reinforcement Learning |