Journal of Pure and Applied MicrobiologyVol. 9 No. Special Edition Nov. 2015

An Enhanced Method for Disease Prediction using Ordinal Classification-APUOC

L. Mary Gladence1, T. Ravi2 and Y. Mistica Dhas3

1Research Scholar, Sathyabama University, Chennai. India. 2Madanapalle Institute of Technology& science, P.B.NO:14, Angallu, Madanapallee - 517 325, India. 3Department of Information Technology, Sathyabama University, Chennai, India.

Received on 26 July 2015 and accepted on 01 October 2015

 

ABSTRACT

Medical Management collects vast amount of data in which retrieving the data in the form of useful as well as not useful data becomes important task. Discovery of this hidden information, Patterns, Relationship often goes undeveloped.Most influential concept in the Data Mining is Classification which has become the most significant method while predicting Disease. Existing work is done based on Nominal Classification which has produced the results in the form of Yes/No.This research uses Ordinal classification concept to produce accurate results. A Novel Technique which proposed in this work can remedy the drawbacks which are faced during earlier research. Proposed work developed a new technique called APUOC. Here, data set is classified based on random threshold value???. Based on these the distance between the different class labels are predicted by All PairsDistance Calculation using Ordinal Classification technique. By doing in this manner there won?t be mismatch while predicting the disease. Taking all these into account, new trained data is in the form of outcome of the proposed work. Here, with the knowledge of proposed work testing data is tested with new trained data sets and the results are predicted. This proposed work is analyzed with existing algorithm such as Kernel Discriminant Analysis, Logistic regression, Classification via Regression, Multiclass classifier etc.,

Keywords : Data collection; Ordinal Classification; Distance Calculation; Projection; Disease Prediction;Data Mining.