In this paper, we attempt to improve the performance of Web proxy cache replacement policies such as LRU and GDSF by adapting a semi naïve Bayesian learning technique. In the first part, Tree Augmented Naive Bayes classifier (TANB) to classify the web log data and predict the classes of web objects to be revisited again future or not. In the second part, a Tree Augmented Naïve Bayes classifier is incorporated with proxy caching policies to form novel approaches known as TANB-LRU and TANB-GDSF. This proposed approach improves the performances of LRU and GDSF in terms of hit and byte hit ratio respectively.
Medical data classification analysis the medical data of the patients to predict the diseases risk. Data mining techniques were highly used in the medical data classification and predicted the diseases. Many existing methods were use the various classifier and feature selection to improve the performance of the classification. Although data imbalance problem is need to be solved for increases the performance. In this research, Synthetic Minority Over-sampling TEchnique (SMOTE) techniques is used for solving the data imbalance problem and Recurrent Neural Network (RNN) was used for the classification. The SMOTE method based on the k Nearest Neighbor (kNN) for the over-sample and under-sample the attributes. The RNN process the instance independent of the previous instance for the classification. Four medical datasets of University of California, Irvine (UCI) were used to evaluate the effectiveness of the proposed SMOTE-RNN method. The proposed SMOTE-RNN method has the accuracy of 85 % while existing method has 82 % accuracy.
R-Tree is a multidimensional indexing structure that forms basis for all the multidimensional indexing structures based on data partitioning. A number of attempts have been made in the past to improve the performance of R-Tree by manipulating the tree parameters and the data parameters. But hardly any attempt had been made to use external parameters such as disk parameters to enhance the performance. This work attempts to improve the performance of R-Tree by efficiently clustering the nodes into input-output units of the hard disk with in the constraint that the independence between the logical and physical organization of the R-Tree should be preserved. Moreover, to preserve the structural and functional properties of R-Tree at any point in the process of clustering, this paper introduces a concept called controlled duplication. Extensive experiments were conducted and the results are tabulated. The improvements are significant and open more avenues for exploration
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.