Skip to main content

Table 2 Comparison of attention prediction accuracies of tested machine learning methods

From: Predicting students’ attention in the classroom from Kinect facial and body features

  Dataset
Method m A A’ B B’ C
1 (Simple tree) 0.694 0.740 0.756 0.771 0.708
2 (Medium tree) 0.638 0.712 0.636 0.674 0.717
3 (Coarse KNN) 0.654 0,701 0.717 0.731 0.734
4 (Medium KNN) 0.633 0.715 0.601 0.640 0.723
5 (Weight. KNN) 0.638 0.721 0.613 0.641 0.725
6 (Bagged trees) 0.655 0.734 0.706 0.747 0.753
7 (Subspace KNN) 0.615 0.703 0.680 0.712 0.738