Everything is written in Python 3, using Pandas and NumPy.
Start by cloning this repository:
git clone https://github.com/iibrahimli/ai_final_project.git
and cd into the project directory
cd ai_final_project
There are 2 libraries (mlp and dtree) and 2 main files (main_mlp.py and main_dtree.py).
To test the MLP:
python3 main_mlp.py
To plot the loss and metrics during training change False to True in the last section of the code. Other network parameters can be tuned too. Custom activation, loss, and metric functions can be implemented in the corresponding .py file (don't forget to inherit from the provided base classes).
To test the decision tree, run
python3 main_dtree.py
Parameters like tree depth, n_groups may be tuned.
| Imran Ibrahimli | Sanan Najafov | Sabina Hajimuradova |
|---|