strtsmrt is a machine learning system that predicts stock prices based on news articles.
##How it works:
- Gets sentiment analysis data on news associated to stocks in the S&P 500 (using Google's Natural Language Cloud API)
- Performs gradient descent on a neural network learning model using TensorFlow
- Uses historical stock data fetched from Yahoo! Finance for training and testing purposes
##Results: Using a day's news sources, strtsmrt predicts the next day's stock price with an average percentage error of 0.5349% using the neural network model with two hidden layers.