Efficient and Scalable Multiple Class Classification using Bee Colony based Probabilistic Approach

Designed for multi-relational explore and learn about important device data classification, and can be widely used in many fields. New classification algorithm Union, naive Bayes, which is the main function of what is known in the literature for the application of multiple classification Union relational environment. The results showed that naive Bayes achieves greater accuracy compared to existing multi-relational algorithm. In addition, the rules of naive Bayes Over draft has a comprehensive database of more properties. There are many possible extensions Baye naive. Currently, naive patterns Baye and confidence to discover LCR repeated use and generation of classification rules. You can find the most important features of each category label using the procedures relating to the extension of the existing framework. Moreover, the current algorithm can improve in terms of improving the efficiency of techning. Relational multiple of the classification algorithm modified by optimization of bee colonies and Naïve Bayes classification rate and a better comparison, Baye Naive. In the process of Bee colony are the complexity increases calculation time complexity also increases. our overall proposal of test data was algorithm. In this dataset, the clearance rate was 92% .Also use another data set (data set abalone) and estimate some little difference in the clearance rate was 91%.