文章基本信息

标题：Cuckoo Search based Hybrid Classification Techniques
本地全文：下载
作者：Ruchika Singh Rajput ; Dr. Jitendra Agrawal ; Dr. Sanjeev Sharma 等
期刊名称：International Journal of Computer Science & Technology
印刷版ISSN：2229-4333
电子版ISSN：0976-8491
出版年度：2017
卷号：8
期号：1
页码：53-57
语种：English
出版社：Ayushmaan Technologies
摘要：Data classification is one of the major tasks in data mining that organizes data in the proper manner to provide enhanced functionality to extract useful information from that data. There are various supervised and unsupervised machine learning techniques like FNN (Fuzzy Neural Network) presented by the researchers to provide an enhanced classification of the dataset. But the performance of the classification highly depends on the selection of the parameters, which is used to classify the dataset. Enhanced subset of parameters can provide enhanced classifiers to classify data. There are various optimization techniques like ACO (Ant Colony Optimization) and some others which are used to provide optimized parameters to classify data. But Cuckoo search is an optimization technique which provides a simple and easy functionality to optimize parameter rather than the other techniques. A review over the various hybrid classification techniques which are used to classify and also uses Cuckoo Search based parameter optimization technique, is presented in this paper. It shows that Cuckoo Search provides enhanced and easy tune with other techniques to enhance performance of the classification. A BDT-SVM and Cuckoo Search based technique is presented for the future to provide enhanced classification for the data.
关键词：Classification;Cuckoo Search;Machine Learning;Fuzzy neural Network.