To search, Click below search items.


All Published Papers Search Service


An Improvement Approach for Reducing Dimensionality of Data with Matrix Decomposition in Data Mining


Sasan Jamshidzadeh, Javad Hosseinkhani


Vol. 16  No. 12  pp. 11-14


Many approaches have been developed for dimensionality reduction. These approaches can broadly be categorized into supervised and unsupervised methods. In case of supervised dimensionality reduction, for any input vector the target value is known, which can be a class label also. In a supervised approach, objective is to select a subset of features that has adequate accuracy to predict the target value. Since there is no correct answer in case of an unsupervised approach, the feature selection cannot be accurately stablished. Most of the unsupervised approaches are designed based on finding a subset of features and variables, in such a way that the part of the major structure and configuration of the dataset is preserved. This thesis studied Singular Value Decomposition (SVD) and its application in the unsupervised feature selection. Modification have been made on this method has been analyzed and evaluated. The result of experiment conducted on real data shows the good performance of the feature selection methods based on Singular Value Decomposition.


Dimensionality of Data, Data Mining, Matrix Decomposition dimensionality reduction