mingboiz / bt4222

Predicting the Stock Market using Twitter NLP and Sentiment Analysis

BT4222 - Applied Machine Learning

Group 5 - Predicting the Stock Market using Twitter NLP and Sentiment Analysis

Contributors

1. Clara
2. Glenn
3. Brian
4. Hui Lin
5. Yang Ming

README

• Demo folder contains our final code for modelling
• Demo folder also contains our deliverable demo showing how to utilise the code for your own needs
• scrape_twitter is our code for scraping Twitter data

Setup

Please ensure all other files in the folder are in the same directory as demo, and the environment is setup with preferably environments.yml or requirements.txt. Crucially, Theano backend for Keras must be enabled.

1. Create a virtual environment within your project directory. Setup the environment and activate it in your terminal:

Conda with environments.yml (preferred)

conda env create -f environment.yml
conda activate bt4222

Pip with requirements.txt

python3 -m venv /path/to/new/virtual/environment
pip install -r requirements.txt
$ C:\Users\...\project_folder> venv\Scripts\activate

2. Clone repository:

git clone https://github.com/mingboi95/bt4222.git