Using TensorFlow machine learning and Twitter sentiment analysis to predict stock trends
"stock-predict.py" runs through these steps:
- Requests an input of a NASDAQ stock quote.
- Downloads the last year's historical data for that stock using Google's resources.
- Trains a neural network that the historical data to predict the stock's closing price for tomorrow.
- Uses Tweepy to find a certain number of tweets about that stock and uses TextBlob to determine if there is a positive or negative trends, using sentiment analysis.
- numpy
- scipy
- pyyaml
- tensorflow
- tweepy
- keras
- requests
- textblob
- Build neural network to classify past year's historical data
- Perform sentiment analysis on collected Tweets
- Create portfolio-builder that allows for multiple stocks to be analyzed
- Build interactive webpage to display data and predictions