Inspiration
As an individual deeply interested in the space of finance and the stock market, I wanted to know what the best way would be to make my money grow. Putting money in the stock market is very effective although can be tricky as you never know whether your stocks will go down or up. This was my inspiration as I wanted to create a program that could help me make better decisions with my investments.
What it Does
The program effectively uses data spanning over a decade to train the model and then uses its knowledge to show the past 60 days prediction along with the prediction for the next day so that you can make better decisions.
How it was Built
I built this using some inspiration online along with using some of my own knowledge regarding machine learning models. By researching heavily upon the stock market, I was able to understand the results I was getting from my program and what issues there were to fix in the future.
Challenges
I ran into an issue where my model wasn't running due to the lack of packages installed and some other bugs. Installing these packages on a MacBook proved to be difficult so I had to troubleshoot on a Windows laptop to help fix things. The code was then brought to my MacBook where installing the packages and running it made everything work smoothly.
Accomplishments that I'm Proud Of
I am glad that I was able to dabble in a bit of machine learning with this project as it is my passion right now. Wanting to step into the field of computer science and engineering, this program proved to fuel the passion. With my desire to also work closely with business, this project helped me research more regarding that. All in all, I feel smarter than when I started the project which is the most important thing.
What I Learned
I learnt a lot about the stock market and ways to predict it. I also learnt lots about the types of algorithms and models that would be the most effective in predicting how the market functions over a period of time.
What's Next
With this project, I hope to add more functionality by increasing the number of days that you can predict for while retaining the accuracy of the model. I also want the user to be able to select their desired company and look at the daily price per day in a visually pleasing manner.
Log in or sign up for Devpost to join the conversation.