Entropy (Data Science) 편집하기 (부분)

==Comparison with Gini Impurity==
Entropy and Gini impurity are similar in purpose but differ in approach:
*'''Calculation Complexity''': Entropy involves logarithmic calculations, which are computationally more intensive than the quadratic calculations used in Gini impurity.
*'''Sensitivity to Class Distribution''': Entropy is more sensitive to changes in class distribution due to the log function, which can make it more precise in some cases.
*'''Bias in Split Selection''': Gini impurity tends to prefer splits that prioritize the majority class, while entropy with information gain may yield more balanced splits.