We know that stock markets across the globe affect one another on a short-term and long-term basis. We wanted to see how we could use machine learning to see how they affect each other, and if we can predict the price of one market based on the open and closing prices of other markets.
A basic Flask App is set up in the Jupyter Notebook provided, where we have developed a simple multiple linear regression using the statsmodel package. For an easy interace, we decided to use React as the frontend. Clone the repo, use npm install
to install all the React dependencies, and then execute npm start
to start the React App. Add some values in each of the fields and hit the predict button to get the predicted value.
The current dataset used in training the model is considerably old; it contains data from 2008 to 2018. In turn, the output values are biased towards the older values. In the future, we can aim to use a more modern, updated dataset to obtain better predictions