StockAI

• 

  Main Page

• 

  Data showing predictions and actual results for each day.

• 

  Graph showcasing the trend of the stock using historical data.

Inspiration

Predicting whether to sell or buy a certain stock/crypto can be hard and confusing so having an application that can take historical data and make predictions can help a user decide what their next move to be.

What it does

The main responsibility of StockAI is to run historical data of a stock of the user's choosing and make predictions for the next day.

How we built it

The main language we used was Python. Within Python, we used a variety of libraries and APIs such as: Pandas Numpy Yahoo-Finance SciKit-Learn Matplotlib Tkinter

The Yahoo-Finance had a crucial role in this project as it was our way to access stock market data. The library holds data for S&P 500 stocks (Apple, Microsoft, Bitcoin, etc.) which are 500 of the biggest companies listed in stock exchanges.

Challenges we ran into

Our weakness was our lack of experience with front-end development and UI design in general. We decided to make a GUI in Python just to show how everything worked. The design seems a little bland which we wish we had more time for.

The other challenge we faced was training the model. It took a lot of trial and error as well as taking a long period of time to process all the data. According to time constraints with the event, the accuracy of the predictions may have errors.

Accomplishments that we're proud of

We are super proud that we learned how to use and build a machine-learning model and applied it to something practical.

What we learned

What's next for StockAI

The next steps would be to continue to train the model so it can be more accurate over time as well as make adjustments to the graphical user interface. We can also make adjustments so it can process data quickly.

Built With

• matplotlib
• numpy
• pandas
• python
• scikit-learn
• tkinter
• yahoo-finance