الفهرس 
IntroductionOnly 14 pages are availabe for public view
 |  | from | 44 |  |  |
| | | from | 44 | | |
Abstract
Data mining or knowledge discovery is the process of analyzing large data in order to extract useful information. In another words, it is the process of collection and exploration of data sets and building models through huge data stores to discover previously unknown outline. In medical field, the process of analyzing data and extracting required knowledge from the data set requires advanced techniques, because these data are very essential for the medical decision. In the dataset, some attributes may be redundant; one can find a reduced set of attributes by removing superfluous attributes, without losing the classification power based on feature reduction process. Feature reduction or attribute reduction is an important step in data mining and classification tasks. It aims at selecting a subset of important and discriminative features. In the classification process, the dimensionality reduction removes the feature that is unnecessary, irrelevant or unimportant, which leads to speeding up learning concept and improves the quality of classification. In artificial intelligence, there are several methods for classification, such as support vector machine (SVM), decision tree, Naïve Bayes and others. In this work, reinforcement learning approach was applied to the classification process based on Q-learning technique. However, there is a small number of researches that use reinforcement learning in classification tasks.