GitHub - charlesbueso/svmStockSentiment: Simple support vector classifier implemented with scikit-learn for classifying positive or negative sentiment associated with the Dow Jones index, using binary label to predict whether the market rose or fell on any given specific date.

Branches Tags

Name		Name	Last commit message	Last commit date
Latest commit History 7 Commits
.ipynb_checkpoints		.ipynb_checkpoints
Combined_News_DJIA.csv		Combined_News_DJIA.csv
README.txt		README.txt
stockSentimentSVM.py		stockSentimentSVM.py
stopWords.txt		stopWords.txt

Repository files navigation

----------------------- R E A D M E ----------------------------

------------- BEFORE RUNNING:
To run, the following text files need to be on source folder of the project:
       - Combined_News_DJIA.csv
       - stopWords.txt

------------- REFERENCES:
Sun, J. (2016, August). Daily News for Stock Market Prediction, Version 1. 
Retrieved [March 9th, 2022] from https://www.kaggle.com/aaron7sun/stocknews.


Stop words extracted from:
    https://www.ranks.nl/stopwords

-------------- NOTES:

Data (news headlines and DJIA 1 or 0 feature) was split as follows:
1989 (70% training, 15% development, 15% testing)
1392 training
298 development
299 testing

----------------------- R E A D M E ----------------------------

About

Simple support vector classifier implemented with scikit-learn for classifying positive or negative sentiment associated with the Dow Jones index, using binary label to predict whether the market rose or fell on any given specific date.

python scikit-learn ml stocks svm-classifier

Readme