INF264-project2

Setup

python version: 3.11

1. Create and activate a virtual environment
2. Install packages with: pip install -r requirements.txt
3. Make sure dataset files are present in the project root folder

Files

• DataExploration.py: Class with methods for visualizing and exploring data
• DataPreprocessing.py: Class with methods for processing data
• ModelOptimization.py: Class with methods for training and validating different models
• notebook.ipynb: Jupyter notebook with the whole process
• models/: folder with pretrained models
• train_and_save.py: trains and saves models in models/ folder
• requirements.txt: required packages for this project
• report.md: markdown file used to generate pdf for final report
• images/: images for report

How to run

The models are pretrained and included in the project, but if you want to retrain the models, run the train_and_save.py file.

All the results can be replicated by running the notebook.