Abstrato

Analysis of liver and diabetes datasets by using unsupervised two-phase neural network techniques

KG Nandha Kumar, T Christopher

Data classification is a vital task in the field of data mining and analytics. In the recent years, big data has become an emerging field of research and it has wide range of research opportunities. This paper represents three unsupervised and novel neural network techniques: Two-phase neural network (TPNN), stack of TPNN (sTPNN), and ensemble of TPNN (eTPNN) for classification of liver and disbetes data. In this study, Diabetes and Liver data are analyzed by using proposed techniques. Bench mark data sets of liver disorder and diabetes patients records are taken from UCI repository and processed by artificial neural networks towards classification of existence of disease. They are also used for the evaluation of proposed techniques. Performance analysis of three neural classification techniques is done by using metrics such as accuracy, precision, recall and F-measure. sTPNN and eTPNN are found better in overall performance in classifying the disease.