Mean, variance, and probabilistic criteria in finite Markov decision processes: a review
Journal of Optimization Theory and Applications
Proceedings of the seventh international conference (1990) on Machine learning
Technical Note: \cal Q-Learning
Machine Learning
Exploration bonuses and dual control
Machine Learning
A near-optimal polynomial time algorithm for learning in certain classes of stochastic games
Artificial Intelligence
Introduction to Reinforcement Learning
Introduction to Reinforcement Learning
Metalearning and neuromodulation
Neural Networks - Computational models of neuromodulation
Control of exploitation-exploration meta-parameter in reinforcement learning
Neural Networks - Computational models of neuromodulation
Learning to Predict by the Methods of Temporal Differences
Machine Learning
Near-Optimal Reinforcement Learning in Polynominal Time
ICML '98 Proceedings of the Fifteenth International Conference on Machine Learning
Hi-index | 0.00 |
This article proposes an adaptive action-selection method for a model-free reinforcement learning system, based on the concept of the 'reliability of internal prediction/estimation'. This concept is realized using an internal variable, called the Reliability Index (RI), which estimates the accuracy of the internal estimator. We define this index for a value function of a temporal difference learning system and substitute it for the temperature parameter of the Boltzmann action-selection rule. Accordingly, the weight of exploratory actions adaptively changes depending on the uncertainty of the prediction. We use this idea for tabular and weighted-sum type value functions. Moreover, we use the RI to adjust the learning coefficient in addition to the temperature parameter, meaning that the reliability becomes a general basis for meta-learning Numerical experiments were performed to examine the behavior of the proposed method. The RI-based Q-learning system demonstrated its features when the adaptive learning coefficient and large RI-discount rate (which indicate how the RI values of future states are reflected in the RI value of the current state) were introduced. Statistical tests confirmed that the algorithm spent more time exploring in the initial phase of learning, but accelerated learning from the midpoint of learning. It is also shown that the proposed method does not work well with the actor-critic models. The limitations of the proposed method and its relationship to relevant research are discussed.