AIAMD

This repository is the code repository for the neural network part of the paper "Application of Artificial Neural Network in AMD Method".

Requirements

• python >= 3.9
• torch >= 1.10.0
• pandas >= 1.3.4
• numpy >= 1.20.3
• scikit-learn >= 0.24.2

Configuration and Running

There are already two trained model files in the repository. The model file in the Diagonal folder is used to predict the diagonal terms of the 8B nuclear Hamiltonian matrix and norm kernel, and the model file in the off-Diagonal folder is used to predict the off-diagonal terms of the 8B nuclear Hamiltonian matrix and norm kernel.

Training data and traditional method result data

This repository does not contain training data and the matrix data of traditional method. These data can be downloaded through Google drive.