Multiple Linear Regression Models in Outlier Detection

To view this paper, you need Flash 9
You have an old version of Flash. Click here to download the latest version.
If this doesn't work, please make sure that JavaScript is enabled on your browser.

Added By: White Globe, , on March 18, 2012

Authors: S.M.A.Khaleelur Rahman, M.Mohamed Sathik,K.Senthamarai Kannan

Cite:

Description: Identifying anomalous values in the real-world database is important both for improving the quality of original data and for reducing the impact of anomalous values in the process of knowledge discovery in databases. Such anomalous values give useful information to the data analyst in discovering useful patterns. Through isolation, these data may be separated and analyzed. The analysis of outliers and influential points is an important step of the regression diagnostics. In this paper, our aim is to detect the points which are very different from the others points. They do not seem to belong to a particular population and behave differently. If these influential points are to be removed it will lead to a different model. Distinction between these points is not always obvious and clear. Hence several indicators are used for identifying and analyzing outliers. Existing methods of outlier detection are based on manual inspection of graphically represented data. In this paper, we present a new approach in automating the process of detecting and isolating outliers. Impact of anomalous values on the dataset has been established by using two indicators DFFITS and Cook’sD. The process is based on modeling the human perception of exceptional values by using multiple linear regression analysis.

Comments /

Add a comment (You can embed videos too, from any video site)

186 Views

0 Favorites

Paper Type: Unknown

Paper key (): 591A2