Generalization over Environments in Reinforcement Learning
نویسندگان
چکیده
We give a method to optimize single-agent behavior for several environments and reinforcement functions by learning in several environments simultaneously in [5]. Now we address the problem of learning in one and applying the policy obtained to other environments. We discuss the influence of the environment on the ability to generalize over other environments. How do good learning environments look like?
منابع مشابه
Generalizing Adversarial Reinforcement Learning
Reinforcement Learning has been used for a number of years in single agent environments. This article reports on our investigation of Reinforcement Learning techniques in a multi-agent and adversarial environment with continuous observable state information. Our framework for evaluating algorithms is two-player hexagonal grid soccer. We introduce an extension to Prioritized Sweeping that allows...
متن کاملTask Space Tile Coding: In-Task and Cross-Task Generalization in Reinforcement Learning
Exploiting the structure of a domain is an important prerequisite for being able to efficiently use reinforcement learning in larger state spaces. In this paper, we show how to benefit from the explicit representation of structural features in so-called structure space aspectualizable state spaces. We introduce task space tile coding as a mechanism to achieve generalization over states with ide...
متن کاملImproved Empirical Methods in Reinforcement-learning Evaluation
OF THE DISSERTATION IMPROVED EMPIRICAL METHODS IN REINFORCEMENT-LEARNING EVALUATION by VUKOSI N. MARIVATE Dissertation Director: Michael L. Littman The central question addressed in this research is ”can we define evaluation methodologies that encourage reinforcement-learning (RL) algorithms to work effectively with real-life data?” First, we address the problem of overfitting. RL algorithms ar...
متن کاملState Generalization with Support Vector Machines in Reinforcement Learning
The conventional reinforcement learning assumes discrete state space. Therefore, it is necessary to make states discrete manually in order to handle continuous state environments. However, if the simple discretization is applied, the number of states increases exponentially with the dimension of the state space, and the learning time increase. In this paper, we propose a state generalization th...
متن کاملCan Deep Reinforcement Learning Solve Erdos-selfridge-spencer Games?
Deep reinforcement learning has achieved many recent successes, but our understanding of its strengths and limitations is hampered by the lack of rich environments in which we can fully characterize optimal behavior, and correspondingly diagnose individual actions against such a characterization. Here we consider a family of combinatorial games, arising from work of Erdos, Selfridge, and Spence...
متن کاملذخیره در منابع من
با ذخیره ی این منبع در منابع من، دسترسی به آن را برای استفاده های بعدی آسان تر کنید
عنوان ژورنال:
- Inteligencia Artificial, Revista Iberoamericana de Inteligencia Artificial
دوره 7 شماره
صفحات -
تاریخ انتشار 2003