Causal Generative Modeling on 11 synthetic datasets Mixed type variables (test)
0.58RF Observed AccuracyDBCM
Evaluation results
| Method | Links | |||||
|---|---|---|---|---|---|---|
| DBCMNumber of training samples=10002026.03 | 0.58 | 3.431 | 0.57 | 12.539 | 0.001 | |
| ANMNumber of training samples=10002026.03 | 0.54 | 1.22 | 0.56 | 6.225 | 0.042 | |
| CFlowNumber of training samples=10002026.03 | 0.53 | 3.161 | 0.62 | 7.184 | 0.001 | |
| KANNumber of training samples=10002026.03 | 0.52 | 1.64 | 0.54 | 9.117 | 0.042 | |
| KAAMNumber of training samples=10002026.03 | 0.52 | 1.89 | 0.54 | 8.405 | 0.024 | |
| KANmixNumber of training samples=10002026.03 | 0.52 | 1.79 | 0.52 | 8.249 | 0.69 | |
| KAAMmixNumber of training samples=10002026.03 | 0.52 | 1.6 | 0.52 | 6.294 | 0.69 |