the aim of this paper is to discuss the concept of the evolutionary and machine learning strategies. The genetic Algorithm is a satisfactory procedure for training multi-layer Feed Forward Neural Networks. The MLFF neural network is a one way signal made of multi layers which has the best trade-off between speed and accuracy. It also deals with pattern classification problems. The RSM predicts the mass transfer parameters which are compared with capabilities of ANN model. To assist the network to perform efficiently the learning mechanisms DL and ELM have been implemented for knowledge accuracy. This paper summarizes Multi-layer feed forward (MFNN) using recent evolutionary Algorithms (GA), Response surface methodology (RSM) for predicting the actual capability of the network and to assist the dataset modulation learning mechanisms like Deep Learning (DL), Extreme Learning Machine(ELM) were used.