tanaymeh / depression-sentiment-prediction

Depression Sentiment Prediction Project

Depression Sentiment Prediction

Predicting Depression from User tweets.

Repository Structure

1. The Folder Code contains all the code in this project.
2. The Folder models contains all the final Models that were trained on data (except one, which is Random Forest Classification model, reason being it's huge size which made it impossible to upload on Github).
3. The File Proposal.pdf contains the original proposal for the project.
4. The File Report.pdf contains the final project report for the project.

Dependencies

To run the code present in this project, you have to have following libraries with their specified versions.

numpy >= 1.15.4
pandas >= 0.23.4
matplotlib >= 3.0.2
wordcloud >= 1.5.0
nltk >= 3.4
sklearn >= 0.20.1
xgboost >= 0.90

The Project has several dependencies (as seen above) which can be easily installed issuing the following command from this directory in your local machine / cloud VM:

$ pip install -r requirements.txt

If that doesn't work for some reason, you can manually install them using;

$ pip install numpy pandas matplotlib sklearn xgboost nltk

Dataset

The Original dataset IS NOT included here due it's massive size. It can be downloaded from here. Note: To download the dataset, one must have a verified Kaggle account.