A Machine Learning project performing sentiment analysis on stock market headlines
This project will use natural language processing to create a sentiment index for stocks in the S&P 500. We are looking at using either a SVG model or a naive bayes model that will be tuned with grid search. We will preprocess the data both to get word stems, and to remove filler words.
We will then make a flask API (and maybe MongoDB) to serve the data, and will render it using JavaScript and D3 (?)
Data Sourced from Kaggle: https://www.kaggle.com/miguelaenlle/massive-stock-news-analysis-db-for-nlpbacktests?select=raw_partner_headlines.csv https://www.kaggle.com/dgawlik/nyse?select=prices.csv
Link to our Tableau Story: https://public.tableau.com/app/profile/zachary.owen/viz/StockMarketSentimentAnalysis/StockMarketAnalysis?publish=yes