Table 3 Comparison of ML models for TQC

From: Faithful novel machine learning for predicting quantum properties

Model

Basic accuracy

Advanced accuracy

baseline

0.49

0.49

GBT baseline22

0.81

0.76

NNN

0.72

0.65

CGNN

0.83

0.80

CCNN

0.76

0.71

CANN

0.80

0.75

  1. Importantly, the CANN and CCNN architectures perform well in comparison to an optimized CGNN architecture. In tests without internal skip connections, these alternative architectures exceeded CGNN performance.
Back to article page