A Hybrid Model in Prediction of Adhd Using Artificial Neural Networks

In this paper, a hybrid artificial network model called DIAGADHD is proposed for the diagnosis of ADHD (Attention Deficit/ Hyperactivity Disorder) using neuro fuzzy technique. This model is a combination of unsupervised training algorithm using self organizing maps and supervised training algorithm using radial basis function. The linguistic values of suspected children are received from the parents or the teachers and then converted into fuzzy membership values. Those values are given as input to the hybrid model and trained for diagnosing ADHD. The approach proposed in this paper uses a hybrid neural network system consisting of Kohonen’s self organizing maps followed by a radial basis function which uses fuzzy membership values as input. The model is trained in two phases on ADHD data. The trained hybrid model is tested for its effective performance and the experimental results are compared with the back propagation algorithm.