American Options ANN

This is still a work in progress, all v0.1.* releases are testing releases.

This is a Python package for American Options Pricing using Artificial Neural Networks (ANN) that assumes the option follows a GARCH process. The package will contain 3 stages of datasets for 3 GARCH models:

1. HN-GARCH
2. Duan-NGARCH
3. GJR GARCH

Project Structure

• ann.py: Contains training and evaluation of the ANN model, as well as the main entry point for the program (ao_ann_main(...))
• loss.py: Contains the function to calculate the different loss measures between target and predicted values.
• model.py: Contains the implementation of the ANN model used for pricing American Options.
• dataset.py: Contains parsing the CSV files and preparing the data for training and testing.
• utils.py: Contains utility functions for the package.

Installation

pip install ao_ann

Running Locally

This project uses the Python package manager uv, this can be installed using the following command:

$ git clone https://github.com/Mustafif/AO_ANN.git
$ cd AO_ANN
$ pip3 install uv # install uv
$ uv sync
$ uv run main.py # run the main.py file

Todo

• Comments in the code
• Documentation
• Example program