DropDAE

Introduction

DropDAE (Dropout Denoising Autoencoder) is a novel deep learning framework that improves imputation for high-dimensional data like single-cell RNA sequencing (scRNA-seq).
It uses:

• A denoising autoencoder structure to reconstruct corrupted input data.
• An optional triplet loss based on consensus clustering to improve separation between similar and dissimilar cells.

Installation

You can install the development version of DropDAE from GitHub:

```r # install.packages(“devtools”) devtools::install_github(“wanlinjuan/DropDAE”)