Stock Market Prediction for Algorithmic Trading using Machine Learning Techniques & Predictive Analytics: An Excel based automated application integrated with R and D3.JS

In this thesis, an attempt has been made to build an automated trading system based on basic Machine Learning algorithms. Based on historical price information, the machine learning models will forecast next day returns of the target stock. A customized trading strategy will then take the model prediction as input and generate actual buy/sell orders and send them to a market simulator where the orders are executed. After training on available data at a particular time interval, this application will back test on out of sample data at a future time interval. The tool extensively uses D3.Js, R, Excel VBA and Phantom JS to provide an integrated appication that not only automatically fetches the data from the web but also develops models, creates interactive hovering data label charts on excel and also test the accuracy of the predictions just at the click of a button! Though the models are very basic, the automation mechanism is effective and can be use by people who still want to rely on basic excel apps for stock data analytics (Tool Demo available here: https://youtu.be/VBx7Ik6aw7c) (Soft Copy of application available here: https://osf.io/6gh8m/ - Scroll down - Archive file available under "Files" heading)