Data Analysis for Information Discovery

Artificial intelligence applications are becoming increasingly popular and producing better results in many areas of research. The quality the depends on quantity data its information content. In recent years, amount available has increased significantly, but this does not always mean more therefore results. aim work is to evaluate effects a new preprocessing method for machine learning. This was designed sparce matrix approximation, it called semi-pivoted QR approximation (SPQR). To best our knowledge, never been applied learning algorithms. works as feature selection algorithm, work, an evaluation performance unsupervised clustering algorithm proposed. obtained compared those using, principal component analysis (PCA). These two methods have various publicly datasets. show that SPQR can achieve comparable using PCA without introducing any transformation original dataset.