Indian Journal of Science and Technology
DOI: 10.17485/ijst/2018/v11i47/132538
Year: 2018, Volume: 11, Issue: 47, Pages: 1-5
Original Article
B. Z. Yahaya *, L. J. Muhammad, N. Abdulganiyyu, F. S. Ishaq and Y. Atomsa
Department of Mathematics and Computer Science, Federal of University Kashere, P.M.B. 0182, Gombe, Nigeria; [email protected], lawan.jibril@fukashere.edu.ng, [email protected], [email protected], [email protected]
*Author for correspondence
B. Z. Yahaya,
Department of Mathematics and Computer Science, Federal of University Kashere, P.M.B. 0182, Gombe, Nigeria; [email protected]
Objective: C4.5 data mining algorithm was scaling up using L’ hospital rule by removing all the logarithms and antilogarithms in its calculation process for mining large dataset. Method/Analysis: L’ hospital rule was employed in order to improve the traditional C4.5 algorithm where the average of the information gain ratio and information gain was used. The Time complexity is a used to determine the efficiency of the improved algorithm over the traditionalC4.5 algorithm. Finding: The study shows that, the improved C4.5 algorithm has the best running time of O(n) compared to traditional C4.5 algorithm which has O(n(log2 n)2 ). Novelty/Improvement: The proposed improved algorithm is more efficient compared to C4.5 algorithm when the dataset is large. Nevertheless, for small dataset is traditional C4.5 is more efficient.
Keywords: Algorithm, C4.5, Classification Technique, Data Mining, Time Complexity
Subscribe now for latest articles and news.