Learning with belief levels

نویسندگان
چکیده

برای دانلود رایگان متن کامل این مقاله و بیش از 32 میلیون مقاله دیگر ابتدا ثبت نام کنید

اگر عضو سایت هستید لطفا وارد حساب کاربری خود شوید

منابع مشابه

Learning with belief levels

We study learning of predicate logics formulas from “elementary facts,” i.e. from the values of the predicates in the given model. Several models of learning are considered, but most of our attention is paid to learning with belief levels. We propose an axiom system which describes what we consider to be a human scientist’s natural behavior when trying to explore these elementary facts. It is p...

متن کامل

Maximum Entropy Learning with Deep Belief Networks

Conventionally, the maximum likelihood (ML) criterion is applied to train a deep belief network (DBN). We present a maximum entropy (ME) learning algorithm for DBNs, designed specifically to handle limited training data. Maximizing only the entropy of parameters in the DBN allows more effective generalization capability, less bias towards data distributions, and robustness to over-fitting compa...

متن کامل

Learning Accurate Belief Nets

Bayesian belief nets (BNs) are typically used to answer a range of queries, where each answer requires computing the probability of a particular hypothesis given some spec-iied evidence. An eeective BN-learning algorithm should, therefore, learn an accurate BN, which returns the correct answers to these speciic queries. This report rst motivates this objective, arguing that it makes eeective us...

متن کامل

Belief Affirming in Learning Processes*

A learning process is belief affirming if the difference between a player's expected payoff in the next period, and the average of his or her past payoffs converges to zero. We show that every smooth discrete fictitious play and every continuous fictitious play is belief affirming. We also provide conditions under which general averaging processes are belief affirming. Journal of Economic Liter...

متن کامل

Parallel Learning of Belief Networks

Learning belief networks from a large dataset over a large domain can be computationally expensive even with a single-link lookahead search. It has been shown that a class of probabilistic domain models cannot be learned correctly by several existing algorithms which employ a single-link lookahead search. When a multi-link lookahead search is used, the computational complexity of the learning a...

متن کامل

ذخیره در منابع من


  با ذخیره ی این منبع در منابع من، دسترسی به آن را برای استفاده های بعدی آسان تر کنید

ژورنال

عنوان ژورنال: Journal of Computer and System Sciences

سال: 2008

ISSN: 0022-0000

DOI: 10.1016/j.jcss.2007.06.007