Stock Price Predictor

FOSL
FOSL

Inspiration

We wanted to develop a neural network that could have applications in financial markets to reduce volatility and to forecast recessions before they happened. After reading headlines on stock market surges and drops, we wanted to provide banks and other financial institutions with the capability to have a picture of what would happen to security prices in the short-term.

What it does

Our model trains to reduce error based off of a testing data set - stock prices connected to the time at which they were recorded - and minimizes error in order to get an accurate model to accurately predict future values.

How we built it

We learned Tensorflow and used mentors available to us to develop a neural network, and went to various workshops such as the Phone2Action intro to AI workshop to get a start on our project.

Challenges we ran into

We had trouble managing our graph as we had little experience with matplotlib in Python, but we soon overcame this obstacle and refined our project.

Accomplishments that we're proud of

We are proud that we were successfully able to complete a stock price predictor that accurately predicts prices for the ticker FOSL in the short-term.

What we learned

We learned how to use Tensorflow and Python and refined our understanding with these software libraries, and gained basic knowledge on financial markets.

What's next for Stock Price Predictor

We hope to detect anomalies in stock prices that indicate that some outside factor has affected a security, such as a bad earnings report for a company, by using calculus and topics like actual error.