Table 3.
Parameters used for Features Selection and Resolution Selection experimentation.
Location | Dataset | DT | kNN | SVM | |||
---|---|---|---|---|---|---|---|
Criterion | Max Depth | Algorithm | Neighbors | c | Gamma | ||
Feature Selection Fitness Gym |
FULL | entropy | 10 | ball_tree | 1 | 10 | 1 |
RFE | entropy | 6 | ball_tree | 1 | 1 | 1 | |
KBEST | entropy | 10 | ball_tree | 1 | 1 | 1 | |
MIN | entropy | 8 | ball_tree | 1 | 1 | 1 | |
Feature Selection Living Room |
FULL | gini | 12 | ball_tree | 1 | N/A | N/A |
RFE | gini | 12 | ball_tree | 1 | N/A | N/A | |
KBEST | gini | 12 | ball_tree | 1 | N/A | N/A | |
MIN | gini | 12 | ball_tree | 1 | N/A | N/A | |
Resolution Selection Fitness Gym |
10 s | entropy | 8 | ball_tree | 1 | 1 | 1 |
10 avg. | entropy | 10 | ball_tree | 1 | 1 | 1 | |
30 s | entropy | 16 | ball_tree | 1 | 1 | 1 | |
30 avg. | entropy | 14 | ball_tree | 1 | 1 | 1 | |
1 min | entropy | 14 | ball_tree | 1 | 1 | 1 | |
1 avg. | gini | 16 | ball_tree | 1 | 1 | 1 | |
Resolution Selection Living Room |
10 s | entropy | 22 | ball_tree | 1 | 100 | 10 |
10 avg. | entropy | 24 | ball_tree | 1 | 100 | 10 | |
30 s | entropy | 22 | ball_tree | 1 | 100 | 10 | |
30 avg. | entropy | 22 | ball_tree | 1 | 100 | 10 | |
1 min | entropy | 20 | ball_tree | 1 | 100 | 10 | |
1 avg. | gini | 20 | ball_tree | 1 | 100 | 10 | |
5 min | entropy | 18 | ball_tree | 1 | 100 | 10 | |
5 avg. | entropy | 12 | ball_tree | 1 | 100 | 10 |