Dr.S.Santhosh Baboo & A.R.J.Vigneswari
There are abundant neural network packages available to train the data sets .In this paper two software namely neurosolutions and matlab is compared to verify the efficiency of prediction of training data sets. The present study deals with the identification of olive ridley sea turtle. This image is initially extracted with its unique features namely, shape, color, length and breadth. These feature extracted image is taken for training and testing for image classification. The performance of training is measured through the time consumption in training the data sets and error rate while training the datasets. The algorithm for this training is feed forward neural network. This algorithm is tested in two different tools and the performance is measured in graphs.