Table 15 Average classification accuracies for comparison algorithms of the proposed method
From: Adaptive multiple imputations of missing values using the class center
Algorithms | Dataset types | ||
|---|---|---|---|
Numerical | Categorical | Mixed | |
Baseline | 64.821 | 67.441 | 61.562 |
+ Checking attribute type (1) | 65.660 | 75.436 | 61.664 |
+ Defining class center (2) | 69.220 | 80.701 | 64.634 |
(1) + (2) | 69.246 | 81.597 | 64.901 |
(1) + (2) + Replacing outlier with STD | 70.857 | 82.161 | 72.041 |
(1) + (2) + Replacing outlier with weight distance and median values (3) | 79.349 | 87.865 | 77.721 |