Table 5.
The Pearson correlation coefficient between the test set’s modality B data and samples imputed from modality A. TDImpute is compared to the DACMVA framework, namely the cross-modal variational auto-encoder (CM-VAE) and CM-VAE with adversarial training (CM-VAE-Adv).
| Missing % | TDImpute | CM-VAE | CM-VAE-Adv |
|---|---|---|---|
| 4 | 0.985 | 0.989 | 0.992 |
| 90 | 0.977 | 0.989 | 0.983 |
| 95 | 0.964 | 0.986 | 0.987 |