for nom only binary splits for interval DT$.$ Consider only multioutcome splits

$+1.1.$ Parent measure $[3],$

$2.$ For each good attribute$/$split alternative ${?}^{\text{s}\text{h}\text{o}\text{w}}$ the following $[30]$

i$.$ measure for each child,

ii$.$ Combined children measure, iii. Gain.

Your der formulation of entropy is incorrect. Need to find. entropy for