Prediction of stock and its analysis deals with the prediction of a company's future stock i.e., prices of the shares of the company through some algorithms and then using the results of the algorithm to maximize profits. If the prediction is similar to that of the actual stock's behavior it could yield a significant amount of profit. There are two different kinds of models used for stock prediction: linear models and Non-Linear model. In this paper, we would be talking about the non-linear models and be implementing non-linear neural networks such as Convolution neural network and Recurrent neural network and make a comparison of the results with the actual results. At last a comparison will be done between the two algorithms stated above to find out which among them was the most effective one and hence could yield a maximum profit when invested in a stock market. Copyright © 2019 American Scientific Publishers All rights reserved.