Skip to main content

Table 1 Clustering comparison of the matrix factorization-based methods in terms of Normalized Mutual information (NMI) and Adjusted Random Index (ARI)

From: A fast and efficient count-based matrix factorization method for detecting cell types from single-cell RNAseq data

Method

Brain

Embryo

Pancreas

 

NMI

ARI

NMI

ARI

NMI

ARI

PCA

0.582

0.339

0.366

0.187

0.630

0.368

Nimfa

0.494

0.258

0.414

0.173

0.456

0.114

NMFEM

0.456

0.264

0.741

0.614

0.435

0.175

tSNE

0.712

0.544

0.658

0.538

0.793

0.652

ZIFA

0.797

0.721

0.888

0.748

0.641

0.429

pCMF

0.787

0.788

0.822

0.659

0.547

0.334

ZINB-WaVE

0.892

0.916

0.888

0.721

0.518

0.342

scNBMF

0.901

0.933

0.908

0.763

0.716

0.472

  1. The number with bold indicates the best performance method and the number with grey represents the second best performance method