In this digitalized world all the data from disparate sources tends to cover the separate aspects of each and every student life that could be stored in many modern university campuses. Moreover, this tends to be challenging one that is to combine the holistic view, predicting the academic performance and to promote the student engagement according to the university. The major motive of this paper is that to predict the more accuracy by means of using Augmented Education. As the initial step of this paper one need to move the data set of college students into the real world and then they may or may not be the online and offline learners that provide the behavioral details inside and outside campus. Most widely used for determining the poor , linear and non linear behavioral changes of the lifestyles that could be estimated by providing the LSTM as the initial step and the second step is that predicting the academic performance. Hence therefore we have used Random Forest provides the best prediction.