a new decision tree using conditional cumulative residual entropy

Decision tree algorithms like id3 necessitate discretization when dealing with continuous values, which can result in loss of information. to tackle this issue, the present article introduces a novel decision tree referred to as conditional cumula- tive residual entropy (ccrdt), speci cally designed for situations where both the target variable and input information are continuous. the ccrdt tree employs conditional cumulative residual entropy as opposed to shannon entropy in the in- formation gain criterion. the article proceeds to compare the outcomes of ccrdt with those of id3 using a genuine data set. additionally, the article deliberates on the notable characteristics of ccrdt, inclusive of its ability to surmount the prob- lem of information loss and facilitate simpler calculations. furthermore, the article puts forth the notion of conditional cumulative residual entropy and its application in the context of decision tree algorithms. ultimately, the article concludes that ccrdt represents a promising approach for decision tree learning involving con- tinuous variables, with potential applications in diverse elds such as data mining, machine learning, and pattern recognition.