Abstract—The introduction of decision trees in the 1980’s
lead on to the thriving research until the 90’s. However,
handling data that are not in a clear form was a problem to be
solved since computers were designed to only calculate crisp
data. This paper presents a method that allows decision trees to
handle fuzzy data. Also, for better accuracy, this paper
introduces decision forests, which is a bundle of decision trees
that decide the case together. Additionally we did an experiment
to prove the performance of new algorithm by comparing it with
support vector machine, which is known as the best algorithm in
data mining field.
Index Terms—Fuzzy decision tree, decision forests, bagging,
entropy, classification, data mining, iris data set.
Jooyeol Yoon and Jun Won Seo are with the Department of Natural
Science, Hankuk Academy of Foreign Studies, Yong-in, Korea (e-mail:
blizzard072@naver.com, jw987@naver.com).
Taeseon Yoon was with Korea University, Seoul, Korea. He is now with
Hankuk Academy of Foreign Studies, Yong-In, Korea (e-mail:
tsyoon@hafs.hs.kr).
[PDF]
Cite: Jooyeol Yun, Jun Won Seo, and Taeseon Yoon, "The New Approach on Fuzzy Decision Forest," Lecture Notes on Software Engineering vol. 4, no. 2, pp. 99-102, 2016.