Conditional Independence Testing on Post-nonlinear synthetic CI benchmark (dz=10, n=1000)
4Type-I Error RateFMCIT
Evaluation results
| Method | Links | ||
|---|---|---|---|
| FMCITscale s=1.00, significance level α=0.05, sample size n=1000, number of trials=1002026.02 | 4 | 1 | |
| CDCITscale s=0.75, significance level α=0.05, sample size n=1000, number of trials=1002026.02 | 4 | 0.74 | |
| CDCITscale s=1.00, significance level α=0.05, sample size n=1000, number of trials=1002026.02 | 4 | 0.62 | |
| FMCITscale s=0.75, significance level α=0.05, sample size n=1000, number of trials=1002026.02 | 5 | 1 | |
| CDCITscale s=0.50, significance level α=0.05, sample size n=1000, number of trials=1002026.02 | 7 | 0.94 | |
| FMCITscale s=0.50, significance level α=0.05, sample size n=1000, number of trials=1002026.02 | 8 | 1 | |
| NNLSCITscale s=0.75, significance level α=0.05, sample size n=1000, number of trials=1002026.02 | 8 | 0.55 | |
| NNLSCITscale s=1.00, significance level α=0.05, sample size n=1000, number of trials=1002026.02 | 8 | 0.32 | |
| NNLSCITscale s=0.50, significance level α=0.05, sample size n=1000, number of trials=1002026.02 | 9 | 0.9 | |
| KCITscale s=1.00, significance level α=0.05, sample size n=1000, number of trials=1002026.02 | 23 | 0.88 | |
| KCITscale s=0.50, significance level α=0.05, sample size n=1000, number of trials=1002026.02 | 24 | 1 | |
| KCITscale s=0.75, significance level α=0.05, sample size n=1000, number of trials=1002026.02 | 24 | 0.97 | |
| EKCIscale s=1.00, significance level α=0.05, sample size n=1000, number of trials=1002026.02 | 44 | 0.91 | |
| EKCIscale s=0.75, significance level α=0.05, sample size n=1000, number of trials=1002026.02 | 45 | 0.98 | |
| EKCIscale s=0.50, significance level α=0.05, sample size n=1000, number of trials=1002026.02 | 47 | 1 |