An Applied Mean Substitutions Technique for Detection of Anomalous Value in Data Mining

Abstract

In the numerical value database, inliers in a database are subset of observations adequately small enough compared to the rest of the observations, which appears to be inconsistent with the remaining data set. They are the result of instant failures or early failures, experienced in many life-test experiments. The problem is how to handle Inliers in a dataset, and how to evaluate the Inliers. This paper describes a revolutionary of approach that uses Inliers detection as a pre-processing step to detect the Inliers and then applies Mean Substitution technique algorithm, hence to analyze the effects of the Inliers on the analysis of dataset.