Learning to Identify Irrelevant State Variables
نویسندگان
چکیده
When they are available, safe state abstractions improve the efficiency of reinforcement learning algorithms by allowing an agent to ignore irrelevant distinctions between states while still learning an optimal policy. Prior work investigated how to incorporate state abstractions into existing algorithms, but most approaches required the user to provide the abstraction. How to discover this kind of domain knowledge automatically remains a challenging open problem. In this paper, we introduce a general approach for testing the validity of a potential state abstraction. We reduce the problem to one of determining whether an action is optimal in every state in a given set. To decide optimality we give two statistical methods, which trade off between computational and sample complexity. One of these methods applies statistical hypothesis testing directly to learned state-action values, and the other applies Monte Carlo sampling to a learned Bayesian model. Finally, we demonstrate the ability of these methods to discriminate between safe and unsafe state abstractions in the familiar Taxi domain.
منابع مشابه
First-Order Bayes-Ball
Efficient probabilistic inference is key to the success of statistical relational learning. One issue that increases the cost of inference is the presence of irrelevant random variables. The Bayes-ball algorithm can identify the requisite variables in a propositional Bayesian network and thus ignore irrelevant variables. This paper presents a lifted version of Bayes-ball, which works directly o...
متن کاملOn the Necessity of Irrelevant Variables
This work explores the effects of relevant and irrelevant boolean variables on the accuracy of classifiers. The analysis uses the assumption that the variables are conditionally independent given the class, and focuses on a natural family of learning algorithms for such sources when the relevant variables have a small advantage over random guessing. The main result is that algorithms relying pr...
متن کاملNon-Verbal Communication in Models of Communicative Competence and L2 Teachers’ Rating
Non-verbal communication (NVC) plays a major role in various aspects of human life (Andersen, 2004; Cameron, 2001; Johnstone, 2008). Children learning their first language come to realize non-verbal communication as their socialization process takes place (Fletcher & German, 1990; Ingram, 1996; Owens, 2001). However, most EFL learners may have little exposure to these non-verbal aspects of comm...
متن کاملExploiting Product Distributions to Identify Relevant Variables of Correlation Immune Functions
A Boolean function f is correlation immune if each input variable is independent of the output, under the uniform distribution on inputs. For example, the parity function is correlation immune. We consider the problem of identifying relevant variables of a correlation immune function, in the presence of irrelevant variables. We address this problem in two different contexts. First, we analyze S...
متن کاملTowards Learning to Ignore Irrelevant State Variables
Hierarchical methods have attracted much recent attention as a means for scaling reinforcement learning algorithms to increasingly complex, real-world tasks. These methods provide two important kinds of abstraction that facilitate learning. First, hierarchies organize actions into temporally abstract high-level tasks. Second, they facilitate task dependent state abstractions that allow each hig...
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:
دوره شماره
صفحات -
تاریخ انتشار 2004