DIRESA - distance-regularized Siamese twin autoencoder

DIRESA-Torch is a Python package for dimension reduction based on PyTorch. The distance-regularized Siamese twin autoencoder architecture is designed to preserve distance (ordering) in latent space while capturing the non-linearities in the datasets.

Introduction:

• DIRESA architecture
• Install DIRESA-Torch

Tutorial:

• DIRESA tutorial

Module reference:

• models module
• trainer module
• loss_funcs module
• latent module
• utils module

Project links:

• Paper
• Code
• Issues