Performance Evaluation Of Machine Learning Algorithms In Dimensionality Reduction

Abstract

The use of data mining has been increasing day to day. This data mining helps to extract the only required data from the large amount of dataset. This is a method of examining the dataset to create new information. Data mining technique is applied for separating the only needed data from large amount of data. The one of the concepts of data mining is PCA which is used to extract the required features. PCA is a process that is used to reduce the number of features by extracting the needed data. After applying this PCA technique in the pre processed data set the features has been overlapped one over the other. Avoiding such kind of problem, cluster the data by using DBSCAN technique.  This clustering method separates the features with huge data to lower density of cluster has more benefits. The concept of DBSCAN algorithm is applied in the database the data is filtered. Here SVM and Naïve Bayes machine learning concepts are applied on the clustered data and find the accuracy level.