Fig. 4
From: Autoencoder-based cluster ensembles for single-cell RNA-seq data analysis
Comparison of autoencoder-based clustering framework with PCA-based dimension reduction and clustering using the four evaluation metrics. Statistical significance (p<0.001; denoted by ⋆) of either autoencoder with k-means clustering against the rest for human and mouse archived brain datasets, or autoencoder with SIMLR clustering against the rest for mouse neurons and striatum datasets were performed using Wilcoxon Rank Sum test (two-sided)