Physiological Signal Analysis and Stress Classification from VR Simulations Using Decision Tree Methods

. 2023 Jun 26;10(7):766. doi: 10.3390/bioengineering10070766

Algorithm 2: Novel Gini index Algorithm.

1 function GiniSplit $(F e a t u r e s, l a b e l s, l e n g t h (X + y))$ ;

Input : features, labels, length(X + y)

Output: gini split, cutoff value, length(X + y)

2 sorted ← $Sort Features$

$l a b e l s_{t o t a l}$ ← $y [sorted]$

$0_{b e l o w c u t o f f}$ ← $find all 0 labels below split point$

$1_{b e l o w c u t o f f}$ ← $find all 1 labels below split point$

$0_{a b o v e c u t o f f}$ ← $total sum of 0 labels -$ $0_{b e l o w c u t o f f}$

$1_{a b o v e c u t o f f}$ ← $total sum of 1 labels -$ $1_{b e l o w c u t o f f}$

$g i n i_{b e l o w}$ ← $multiply ratios of two below labels$

$g i n i_{a b o v e}$ ← $multiply ratios of two above labels$

$I n i t i a l G i n i_{s p l i t}$ ← $add$ $g i n i_{a b o v e}$ and $g i n i_{b e l o w}$

$G i n i_{s p l i t}$ ← $find \min {gini}^{’} s from$ $I n i t i a l G i n i_{s p l i t}$

$c u t o f f_{i n d e x}$ ← ${InitialGini}_{s p l i t} [G i n i_{s p l i t}]$

cutoff value ← $X [{cutoff}_{i n d e x}, l e n g t h (X + y)]$

3 return $G i n i_{s p l i t}$ , cutoff value, $l e n g t h (X + y)$