In our previous project, we worked on providing accurate investment decisions when it comes to Canadian Banks. We have worked on 4 major Canadian Banks, Toronto Dominion Bank, Bank of Montreal, Bank of Nova Scotia and Royal Bank of Canada. We have compared their performance over the last 6 years with respect to S&P500.
In this project, our team has worked on implementing Machine Learning Techniques to predict the performance of those Canadian Banks, taking their historical data into accounts and predict how they might perform in the future.
We have deployed the Machine Learning methods used behind the scene and deployed them into a streamlit app, making it simple and easy for anyone who wants to extract feature, view the performance or even train their own data!
Our model provides business solutions to a vast customer base such as retail traders or anyone who wants to know more about a particular stock before making investment decisions.
Although our model is based on 4 Canadian Banks, because our main focus was to provide easier investment decisions, our this model could be used for any sort of stocks. All the user of this app needs to do is to select the ticker, and choose from our multi-option button what they want to do.
On top of that, our app is also equiped to give the users the opportunity to select the batch and epochs number he wants to train the model with, and fetch results.
Our goal was to take our previous project where we only showed the client which banking stocks he wants to invest in based on his willingness of risk and take it a step further by making it easier for the client to dive into the technical aspects of any sort of investment decisions.
Although this app is at the momonet restricted with only four stocks, it can be also trained with any sort of stocks data or basically any other data the customer demands, and the fetched results can be expected to be as accurate as our model.
Our aim is to make this app more user friendly, adding more features where not only stock investment decisions could be made, but it will also contain more investment areas such as cryptocurrency and even real state. Where customers would choose the choose the cryptocurrency tickers. They would also be able to choose areas where they would want to purchase a property and make comparison between properties where they wants to invest in and see which investment decisions would yeield the maximum profit in the coming days.
Finally, we aim to introduce more features for our model to train in such as NLP to get more accurate results.