High impact event processing using incremetal clustering in unsupervised feature space through genetic algorithm by selective repeat ARQ protocol

High impact event represents the information which are frequently used. The frequently used information is maintained in different clusters such that it can be accessed quickly without involving much searching time. Clustering methods are one of the key steps that lead to the transformation of data to knowledge. Clustering algorithms aims at partitioning an initial set of objects into disjoint groups (clusters) such that objects in the same subset are more similar to each other than objects in different groups. In this paper we present a generalization of the k-Windows clustering algorithm in metric spaces by following a selective Repeat ARQ protocol having fixed window size for accurate information transmission. The original algorithm was designed to work on data with numerical values. The proposed generalization does not assume anything about the nature of the data, but only considers the distance function over the data set. The efficiency of the proposed approach is demonstrated on msnbc data sets. Genetic algorithm approach is used to detect and predict high-impact events in different areas such as automotive manufacturing, networking for data transmission, etc. While the high-impact events occurs infrequently, they are quite costly, means they have high-impact on the system key performance indicator. This approach is based on mining these types of events and its impact on the total process execution. The classified data are clustered for future implementation which have similar feature. Due to the clustering concept the clustered data can be used for various applications, which makes it robust. The parameters are optimized for best solution. This approach is tested on high impact events that occurs in networking, during transmission and it was found to be robust, highly accurate and with less probability of fault, for prediction of future occurrences of such events.