Stockopedia

This project was part of a hackathon hosted by India’s National Stock Exchange. Our problem statement was to come up with a fin-tech product to lead to better trading practices. After initial brainstorming, we came up with the idea of building a stock recommendation system for short-term traders. We ended up being finalists at the hackathon which involved more than 165 participants.

Having taken an online course title 'Python for Finance: Investment Fundamentals and Data Analytics', I was already aware of concepts like rate of return, risk diversification and so on. So we built a product that took in a stock’s score about the rate of return, risk diversification, news sentiments, mutual fund activity and short-term predictions with respect to a client’s existing portfolio to recommend top buy/sell options.

We primarily worked with Python in this project. The short-term predictions were calculated using a LSTM model and the news sentiment score was calculated using logistic regression.

The project files can be found here.