Characterizing Potential Outliers using Confidence Interval Ellipse and Regression bands for Traffic Data Analysis

Characterizing Classes of Potential Outliers through Traffic DataSet Data Signature 2D nMDS Projection*

Erlo Robert F. Oquendo, Jhoirene B. Clemente, Jasmine A. Malinao, Henry N. Adorna
Department of Computer Science (Algorithm and Complexity Laboratory),

University of the PhilippinesVelasquez Ave., Diliman, Quezon City 1101

Email: {efoquendo, jbclemente}, {jamalinao, hnadorna}



This paper presents a formal method for characterizing thepotential outliers from the data signature projection of trafficdata set using Non-Metric Multidimensional Scaling (nMDS)visualization. Previous work had only relied on visual inspectionand the subjective nature of this technique may derive false andinvalid potential outliers. The identification of correct potentialoutliers had already been an open problem proposed in literature.This is due to the fact that they pinpoint areas and time frameswhere traffic incidents/accidents occur along the North LuzonExpressway (NLEX) in Luzon.

In this paper, potential outliers are classified into (1) absolutepotential outliers; (2) valid potential outliers; and (3) ambiguouspotential outliers through the use of confidence bands andconfidence ellipse. A method is also described to validate clustermembership of identified ambiguous potential outliers.

Using the 2006 NLEX Balintawak Northbound (BLK-NB) dataset, we were able to identify two absolute potential outliers, ninevalid potential outliers, and five ambiguous potential outliers.

In a literature where Vector Fusion was used, 10 potentialoutliers were identified. Given the results for the nMDSvisualization using the confidence bands and confidence ellipses,all of these 10 potential outliers were also found and 8 newpotential outliers were also found.

Download Full PDF

This paper was published in the conference proceedings of 2010 National Conference on Information Technology  Education (NCITE) held at La Carmela de Boracay, Boracay, Philippines.


Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s