A PyTorch implementation for Transition Matrix Representation of Trees with Transposed Convolutions

This project is a PyTorch implementation for Transition Matrix Representation of Trees with Transposed Convolutions, published as a conference proceeding at
SDM 2022. The paper proposes TART (Transition Matrix Representation with
Transposed Convolutions), a novel framework for generalizing tree models with a
unifying view.

Requirements

The repository is written by Python 3.7 with the packages listed in
requirements.txt. A GPU environment is strongly recommended for efficient
training and inference of our model. You can type the following command to
install the required packages:

pip install -r requirements.txt

Datasets

The paper uses 121 datasets from the UCI repository. Since the size of all
datasets is larger than 500 MB, we include only balance-scale in the current
repository, which is a sample dataset that includes 625 examples of

To finish reading, please visit source site