This project explores incorporating uncertainty into binary classification of tweets, specifically identifying tweets related to disasters. It experiments with blending traditional NLP techniques and Bayesian inference to provide not only a classification but also a measure of confidence in that classification.

• NLP Model: Utilizes GPT-2 for converting tweets into embeddings, serving as a nuanced representation of the textual data.
• Classification Model: A neural network built with PyTorch that takes GPT-2 embeddings as input and predicts two parameters (alpha and beta) of a Beta distribution. This distribution represents the probability of a tweet being disaster-related along with the uncertainty of the prediction.
  • Input Layer: Receives the GPT-2 embeddings.
  • Hidden Layer(s): Processes the embeddings through one or more layers with non-linear activation functions.
  • Output Layer: Outputs two values, alpha and beta, parameters of the Beta distribution.
• Loss Function: A custom loss function that combines Binary Cross-Entropy (BCE) for accuracy and a penalty for variance, encouraging the model to be both accurate and confident in its predictions.

To get started, clone the repository, set up a virtual environment, install dependencies, and run the provided Jupyter notebook:

git clone <repository-url>
cd <repository-directory>
python -m venv venv
source venv/bin/activate  # On Windows use `venv\Scripts\activate`
pip install -r requirements.txt
jupyter notebook disaster_tweet_classifier.ipynb