Implementing Agents in Classification with Partition Method

E. Padmalatha; Dr. C. R. K. Reddy; Dr. Padmaja Rani

Knowledge management needs very high end accuracy and efficiency in making decisions, this can be achieve by implement agent system with data mining. Heterogeneous multi agent coordination improves accuracy of large and complex data mining task. A working agent is a data mining agent designed to ferret out relevant information to classify the data stream. On one hand, data mining approach can drive the knowledge extraction from huge amount of data; on the other hand, by using a single classifier it may not provide reliable results. Therefore, to improve the classification accuracy rates, ensemble classifiers joint with agents and relied on multiple learning algorithms, can be aid to improve the performance. The main objective of the proposed system is to improve the accuracy of the classification model by using ensemble technique with the data mining agents. Data can be divided into partitions according to the partitions the number of agents will created and the partitions are processed by all the agents in the systems in equivalent, the results from the partitions are fused to consider the accuracy. Finally accuracies are compared with respective to those achieved by using classifier.

Select Volume / Issues:

• Volume 6,Issue 6, Nov. 2015

Year:

2015

Type of Publication:

Article

Keywords:

Agent; Data Stream; Ensemble Classification

Journal:

IJECCE

Volume:

6

Number:

6

Pages:

641-644

Month:

November