Lifelong Hyper-Policy Optimization with Multiple Importance Sampling Regularization
Autor: | Liotet, Pierre, Vidaich, Francesco, Metelli, Alberto Maria, Restelli, Marcello |
---|---|
Rok vydání: | 2021 |
Předmět: | |
Druh dokumentu: | Working Paper |
Popis: | Learning in a lifelong setting, where the dynamics continually evolve, is a hard challenge for current reinforcement learning algorithms. Yet this would be a much needed feature for practical applications. In this paper, we propose an approach which learns a hyper-policy, whose input is time, that outputs the parameters of the policy to be queried at that time. This hyper-policy is trained to maximize the estimated future performance, efficiently reusing past data by means of importance sampling, at the cost of introducing a controlled bias. We combine the future performance estimate with the past performance to mitigate catastrophic forgetting. To avoid overfitting the collected data, we derive a differentiable variance bound that we embed as a penalization term. Finally, we empirically validate our approach, in comparison with state-of-the-art algorithms, on realistic environments, including water resource management and trading. Comment: Accepted at AAAI2022 |
Databáze: | arXiv |
Externí odkaz: |
načítá se...