THE EFFECT OF AMOUNT OF DATA ON RESULTS OF ACCURACY VALUE OF C4.5 ALGORITHM ON STUDENT ACHIEVEMENT INDEX DATA
| Autor: | Anton Sunardi, Sienny Rusli, Christina Juliane |
|---|---|
| Jazyk: | angličtina |
| Rok vydání: | 2022 |
| Předmět: | |
| Zdroj: | Jurnal Riset Informatika, Vol 4, Iss 2, Pp 191-198 (2022) |
| Druh dokumentu: | article |
| ISSN: | 2656-1743 2656-1735 |
| DOI: | 10.34288/jri.v4i2.374 |
| Popis: | Of the many academic data, data in the form of an achievement index needs to be used in-depth so that it does not become a display of numbers and information only. This achievement index evaluation data reflects the educational process students and teaching staff carries out in an educational process. This study aims to measure the accuracy of data mining processing based on differences in test data by analyzing the C4.5 algorithm using RapidMiner as a data processing tool and determining the decisions students can make and academic institutions in developing study strategies and educational curricula to be maximized. The data processing is carried out by classifying the student achievement index data at a private university using data analysis test equipment. The data source comes from kaggle.com, which consists of 1687 data that have been processed and processed. The conclusion from the results of this study is that the amount of data turns out to have a significant influence on the accuracy value of the C4.5 algorithm, where an accuracy rate of 91.69% is obtained from the test results of 1687 data with four main attributes, namely IPK1, IPK2, IPK3, IPK4 and correctly or not as a label. |
| Databáze: | Directory of Open Access Journals |
| Externí odkaz: |
|