Skip to main content
. 2018 Apr 3;13:5. doi: 10.1186/s13062-018-0207-8

Table 3.

Comparison of classification performance on ALL-EFS and ALL-OS endpoints using the juxt AG1-G/CNV-G dataset or its embedding obtained using the autoencoder approach

AG1-G/CNV-G AG1-G/CNV-G encoding
Endpoint MCC CV MCC val Nfeat MCC CV MCC val Nfeat
ALL-EFS 0.207 (0.141-0.278) 0.197 1000 0.438 (0.374-0.503) 0.360 100
ALL-OS 0.523 (0.443-0.591) 0.359 500 0.545 (0.466- 0.618) 0.538 100