Search

High dimensional data have brobdingnagian number of features, but not all features are useful. Irrelevant and redundant features may even reduce the classification accuracy. Feature selection is a process of selecting a subset of relevant features to decrease the dimensionality of data. When applied on high dimensional datasets (Big Data) the feature selection methods perceives many challenges and it is pertinent to come up with the new methods or revamp the existing methods. In this study, a new method 'Newtonian particle swarm optimization (NPSO)' has been proposed. In the proposed method Newton's second law of motion has been used to update the learning mechanism of PSO. In NPSO, particle not only learn from the position but also from the mass and acceleration of neighboring particles. The proposed method is mathematically validated at equilibrium using eigen values. Further, the proposed method has been applied on high dimensional microarray gene expression dataset. The NPSO is also compared with other state of art feature selection methods. Selected features, classification accuracy and dimension reduction are used to appraise the goodness of the proposed method. Mathematical validation and experimental results clearly validates the merits of the proposed method in field of feature selection. This paper show the classwise analysis of SRBCT, Brain1, 11-Tumor and 14-Tumor datasets. When number of classes increased dimension reduction is increased but classification accuracy of dataset is decreased. [ABSTRACT FROM AUTHOR]

Dimensionality reduction of high dimensional data still perceives challenges and hence, it is pertinent to introduce new methods or revamp existing methods. In this study, a new ternary particle swarm optimization (TPSO) algorithm has been proposed, in which particle is a string of "trit", which is the smallest unit of information. Ternary string is made up of (0, 1 and #). 0, 1 and # are representatives of rejection, acceptance and intermediate (uncertain) states respectively. Since trit is the smallest unit of information therefore, # trit brings the characteristics of quantum theory in the search. This provides the better exploration of feature leading to global optimum solution. This method belongs to wrapper category of feature selection method since it has k nearest neighbor classifier as performance evaluator. The TPSO has been applied in two phases. The second phase is included in the system in a top down information processing fashion, in which big system of information is broken down to have insight into the hidden important information. In first phase TPSO is applied multiple times on each data set. In second phase optimum features are retrieved by applying LBUB, EXP SEARCH and VOTE MERGE methods. Experimental results on bench marking datasets show that the proposed methods are promising to handle feature selection in high dimensional space. [ABSTRACT FROM AUTHOR]