Forcasting the stock market movements are usually based on only one data source and thus may not cover factors that can affect the stocks.Here I have exploit the consistencies among differnt data sources and developed a multi-source instance model that can combine sentiments as well as quantative data into comprehensive framework.This approach is able to determine the importance of each data source identify the crucial input information that is considered to drive the movements, making the prediction interpretable.
Stock prediction has been focused on sincea long preiod of time.So what was the means of prediction of stock prices when we don't have sequential models that understand the dependency of the current data on the previous data over the time sequence?
Well before the use of Deep Learning in the field of stock prediction, the process of building robust models for stock price prediction has started.But sequence models changed everything and led to a massive improvement in state-of-the-art results in the finance industry.
The purpose of this project is to predict closing prices of stocks using twitter tweet's from various users and historical data from yahoo finance, using Deep Learning model, Long-Short Term Memory (LSTM) neural network.
If you are further interested in the other projects checkout this link here.
- sys (https://docs.python.org/3/library/sys.html)
- os (https://docs.python.org/3/library/os.html)
- numpy (https://www.numpy.org/)
- tweepy (https://www.tweepy.org/)
- textblob (https://textblob.readthedocs.io/en/dev/)
- yfinance (https://pypi.org/project/yfinance/)
- Tensorflow (https://www.tensorflow.org/)
- keras (https://keras.io/)
Before you go and run the code, Install all dependencies and also make an application on twitter dev page and get your keys for authentication.
Let me know of more updates you'd like to see.