Open AccessImplementation of the C45 Algorithm in Classifying Classes
Author(s) -
Nellysa Putri Denila Nasution,
Fristi Riandari
Publication year - 2022
Publication title -
jurnal mandiri it
Language(s) - English
Resource type - Journals
eISSN - 2809-1884
pISSN - 2301-8984
DOI - 10.35335/mandiri.v10i2.92
Subject(s) - arabic , decision tree , class (philosophy) , information gain ratio , mathematics , computer science , rank (graph theory) , value (mathematics) , process (computing) , algorithm , artificial intelligence , data mining , machine learning , statistics , combinatorics , linguistics , philosophy , operating system
Mts Swasta YPII Kotarih is one of the educational institutions which in its implementation classifies several classes for students, one of which is the superior class. However, in practice, each person is still selected based on the rank in each class, making the classification process less accurate and efficient, a computerized classification technique is needed in its implementation, the classification technique used by Data Mining is the c4.5 algorithm that works with the Decision tree to find the highest gain value from each criterion existing criteria, the criteria used in this study are the subjects of Q. Hadith, Akidah, Fiqh, SKI, Arabic, Tahfizh, General research results show that from There are 3 criteria that are most influential in the classification process, namely General, Arabic, Fiqh, implemented in the form of a computerized system. Based on the research results, the highest gain value is the highest gain is General with a value of 0.3629, so that the general attribute becomes node 1 or the root of the decision tree. second highest gain B.Arabic attribute, with a gain value of 1.7189, and the third highest gain is Fiqh with a value of 0.3486.