Development of User-friendly Stock Prediction Assistance using Ensemble Learning

Taking informed investment decisions is an extremely important aspect of stock investment. Investors rely upon various websites for assisting them in their decisions, but the lack of personalized approach limits the vision provided by such web-assisted tools. Most of the work in this area is arranged around solving the user’s problem either by using intricate machine learning algorithms or by some high-end mathematical or statistical methods.In this paper, we propose a user-friendly personalized GUI which enables a user at any experience level sail through smoothly with the stock investment process. We use a hybrid approach–in which we use simple statistical tools for prediction and provide the ultimate decision using ensemble learning. The statistical methods used are popular and simplistic in nature which make the entire process easy to understand for the user. The ensemble learning model acts as a black-box for the user which will train itself to choose the final decision of "buy", "sell" or "hold" for the user.