Financial crashes, such as the 2008 recession, are often caused by events which are relatively unpredictable. Because of the severe adverse effects of such a depression in the economy, it is essential to find a reliable method to predict the trends in the stock market.
What it does
Our product analyzes Twitter trends and keywords in relation to S&P 500 companies and their stock history to determine how the company will fare in the near future.
How we built it
After searching and collecting data from the Twitter world for mentions of the S&P 500 companies, we input this data into a program which graphs the relations between date, sentimental analysis of tweets, and stock value.
Challenges we ran into
Our main challenge was developing a neural network which connects long short term memory (LSTM) cells and the sentimental analysis data received in the first part.
Accomplishments that we're proud of
We are proud of successfully utilizing LSTM cells to use the historical stock data values from the past as part of our data analysis.
What we learned
We learned a lot in terms of integration between several different parts and languages. Additionally, we gained experience in graphing data and using regression towards predicting a trend.
What's next for Stock Specs
We hope to expand our model making it a viable solution for analysts of many types of companies, including privately traded and international ones