This is a concise Java implementation of maximum entropy, which provides training and prediction interfaces. Training using GIS training algorithm, with sample training set. The purpose of this paper is to introduce the principle, classification and implementation of maximum entropy, which does not involve formula derivation or other training algorithms, please feel free to eat. Introduction to Maximum entropy theory the maximum entropy belongs to the identification model, which satisfies all the known constraints and makes no undue assumptions about the unknown information. What do you call a known constraint? This article does not confuse you with obscure terminology, see an example where your friends "go out" or "self-house" every day, both of which are affected by the "weather", "mood", "humidity" (because she is a girl) and we can call it a feature. Then we collected some "activity <" from her microblog history.
Continue reading : yards Farm» Maximum entropy Java implementation
original link : http://www.hankcs.com/nlp/maximum-entropy-java-implementation.html
Java implementation of maximum entropy
