A Novel Efficiency Enhanced Classifier for Predicting the Attention-Deficit Hyperactivity Disorder

P. Preetha and Dr.R. Mallika

Attention-Deficit/Hyperactivity Disorder (ADHD) refers to a brain disorder indicated by an enduring pattern of inattentiveness and/or hyperactivity along with impulsivity, which interferes with functioning or development. There is not even a distinct test, which can be utilized for the diagnosis of attention lacking hyperactivity disorder observed in children and bigger adults. The diagnosis of ADHD is done just after a person has exhibited few or all the symptoms of ADHD constantly for a period greater than six months in the advanced stage. A psychiatrist has to do more complex and tedious evaluations for diagnosing the ADHD. Moreover, it requires Psychiatrists interpretation and many cases it will lead to mis diagnosis. Therefore, an intelligent automated diagnosis system using machine learning technique is needed for the classification of the ADHD. In this paper an Efficiency Enhanced Extreme Learning is designed for the prediction of the presence of ADHD. Efficiency Enhanced Extreme Learning machine has low covariance and low mean square error than traditional Extreme Learning Machine. In addition, with this it solves the multi collinear problem of Traditional ELM. Performance Comparison is made with Traditional Extreme Learning Machine, Weighted ELM and Support Vector Machine and it is noticed from the results that the newly introduced method outperforms than other classifiers.

Volume 11 | 11-Special Issue

Pages: 418-430

DOI: 10.5373/JARDCS/V11SP11/20193050