Skip to main content

Table 2 The influence of KM, SC, and AG on the ScLSTM model is evaluated on eight data sets

From: ScLSTM: single-cell type detection by siamese recurrent network and hierarchical clustering

 

ARI

NMI

ACC

BAS

KM

SC

AG

KM

SC

AG

KM

SC

AG

KM

SC

AG

Biase

1

0.8825

1

1

0.9008

1

1

0.9464

1

1

0.9321

1

Yan

0.9122

0.672

0.9122

0.9133

0.8259

0.9133

0.8778

0.7111

0.8778

0.965

0.7992

0.965

Goolam

0.9925

0.604

0.9925

0.9651

0.7408

0.9651

0.9758

0.7177

0.9758

0.9965

0.782

0.9965

Camp

0.8197

0.6505

0.8356

0.888

0.8083

0.8984

0.8636

0.8005

0.8841

0.9274

0.8452

0.9462

Kolodziejczyk

1

1

1

1

1

1

1

1

1

1

1

1

Li

0.9641

0.9822

0.9527

0.963

0.9795

0.9573

0.9768

0.9893

0.975

0.9765

0.9886

0.9736

Usoskin

0.9771

0.9771

0.9885

0.9634

0.9634

0.98

0.9904

0.9904

0.9952

0.9896

0.9896

0.994

Pollen

0.9584

0.6471

0.9584

0.9793

0.8574

0.9793

0.9601

0.7143

0.9601

0.9934

0.8544

0.9934

  1. The best results are shown in bold
Back to article page

BMC Bioinformatics

ISSN: 1471-2105

Contact us