Over the past decade, data mining has gained an important role in analysis of large datasets and there by understanding the complex systems in almost all areas. Such datasets are often collected in a geographically distributed way, and cannot, in practice, be gathered in to single repository. Existing data mining methods for distributed data are of communication intensive. Many algorithms for data mining have been proposed for a data at a single location and some at multiple locations with improvement in terms of efficiency of algorithms as a part of quality but effectiveness of these algorithms in real time distributed environment are not addressed, as data on the web/network are distributed by very of its nature. As a consequence, both new architectures and new algorithms are needed. In this paper we introduce the software agent technology that supports building of distributed data mining architecture and explore the capabilities of mobile agents' paradigm and will show by experiment that, it is suited for distributed data Mining compared to traditional approaches like client server computing..

AbstractThe main objective of the research is to provide a multi-agent data mining system for diagnosing diabetes. Here, we use multi-agents for diagnosing diabetes such as user agent, connection agent, updation agent, and security agent, in which each agent performs their own task under the coordination of the connection agent. For secure communication, the user symptoms are encrypted with the help of Elliptic Curve Cryptography and Optimal Advanced Encryption Standard. In Optimal Advanced Encryption Standard algorithm, the key values are optimally selected by means of differential evaluation algorithm. After receiving the encrypted data, the suggested method needs to find the diabetes level of the user through multiple kernel support vector machine algorithm. Based on that, the agent prescribes the drugs for the corresponding user. The performance of the proposed technique is evaluated by classification accuracy, sensitivity, specificity, precision, recall, execution time and memory value. The proposed method will be implemented in JAVA platform.

Abstract:A multi agent distributed data mining system for diagnosing diabetes and classification is proposed. Here we are introducing four agents namely user agent, connection agent, updation agent, and security agent. In which each agent performs their own task under the coordination of the connection agent. The user agent collects the user symptoms in order to predict the patient status also the knowledge based of the system. Updation agent is responsible for prescribing drugs for the patient. For secure communication, the proposed technique introduces one security agent between connection agent and updation agent. Here the user symptoms are encrypted by means of advanced encryption standard (AES). Finally, updation agent is classifying the user symptoms and then evaluates the diabetes level with the help of hybrid firefly based neural network algorithm. The performance of the proposed system will acquire with the classification accuracy. The proposed method will be implemented in JAVA platform.

No abstract

No abstract