The fabulous growth of Intelligence Computing leads to the progress of Human Computer Interaction. The Human Nervous System and behavior patterns interactive with intelligence based machine. Now a days, Lot of researchers are concentrate in this area. In this paper, we analysis the existing methodology and also find the feature selection methods. We propose a new Feature Selection method such as Z-Test, Population Variance, Population Standard Deviation, Sample Variance and Sample Standard Deviation. It will outperformed better than some of the benchmark feature selection Methodology.