This was Team Init To Win It's submission for the Datathon.
File descriptions:
- pre_processing_and_analysis.ipynb: For pre-processing the provided datasets, and then performing EDA and various other analyses
- feature_importance.ipynb: For identifying features, as individuals and in sequence, which were most impactful, using Logistic Regression, Support Vector Machine, Ridge Classifier, and XGBoost.
- radial_viz.ipynb: A Radial graph in contention for best visualization.
- modeling.ipynb: For modeling and subsequent predictions, using LSTM and Transformers. We also provided an LLM-based solution, testing Google's BERT in the process, but settling with OpenAI's GPT-3.5 Turbo API.