Table 5 Comparison of results evaluated on RF for various missing value algorithms
S. No. | Missing value algorithm name | Hyperparameter | Mean accuracy score with RF | Std accuracy score with RF |
|---|---|---|---|---|
1 | Statistical imputation | Mean | 0.67 | 0.06 |
Median | 0.66 | 0.06 | ||
Most_frequent | 0.67 | 0.05 | ||
Constant | 0.68 | 0.06 | ||
2 | KNN imputation algorithm | No. of neighbour 1 | 0.67 | 0.06 |
No. of neighbours 3 | 0.67 | 0.04 | ||
No. of neighbours 5 | 0.66 | 0.07 | ||
No. of neighbours 7 | 0.65 | 0.06 | ||
No. of neighbours 9 | 0.66 | 0.06 | ||
No. of neighbours 15 | 0.66 | 0.05 | ||
No. of neighbours 18 | 0.66 | 0.07 | ||
3 | Iterative imputation | Iterative imputation using Ascending hyperparameter | 0.68 | 0.05 |
Iterative imputation using Descending hyperparameter | 0.67 | 0.07 | ||
Iterative imputation using Roman hyperparameter | 0.67 | 0.05 | ||
Running iterative imputation for 1 iteration | 0.68 | 0.06 | ||
Running iterative imputation for 2 iterations | 0.66 | 0.06 | ||
Running iterative imputation for 3 iterations | 0.67 | 0.06 | ||
Running iterative imputation for 4 iterations | 0.68 | 0.06 | ||
Running iterative imputation for 5 iterations | 0.66 | 0.09 | ||
Running iterative imputation for 6 iterations | 0.66 | 0.06 | ||
Running iterative imputation for 7 iterations | 0.67 | 0.08 |