التفاصيل البيبلوغرافية
| العنوان: |
An Improved Accuracy of Multiclass Random Forest Classifier with Continuous Attribute Transformation Using Random Percentile Generation. |
| المؤلفون: |
Susetyoko, Ronny, Purwantini, Elly, Iman, Budi Nur, Satriyanto, Edi |
| المصدر: |
International Journal on Advanced Science, Engineering & Information Technology; 2023, Vol. 13 Issue 3, p943-953, 11p |
| مصطلحات موضوعية: |
RANDOM forest algorithms, PERCENTILES, DATES (Fruit), TUITION |
| مستخلص: |
This study aims to improve classification accuracy by transforming continuous attributes into categories by randomly generating percentile values as categorization limits. Four algorithms were compared for the generation of percentile values and selected based on the small variability of the percentile values and the distribution of the highest revenue expectations. The distribution of testing and training data classification accuracy becomes the second consideration. Random forest (RF) classification is modeled from selected percentiles with three transformation variations. The results of the ANOVA test, the algorithm with three variations of the transformation, has a mean that is not significantly different from the best model and the original dataset model. However, in some variations of training data, RF classification with continuous attribute transformation was superior to the original dataset model. The effectiveness of this continuous attribute transformation algorithm was very well applied to the LR, MLP, and NB methods. In the tuition fee dataset, the application of the algorithm for the three methods each had an accuracy of 0.178, 0.204, and 0.318. The results of the attribute transformation give a significant increase in accuracy to 0.967, 0.949, and 0.594 for each method, respectively. In the date fruits dataset, the attribute transformation was effective in the MLP method with an accuracy of 0.193 (original attribute) to 0.690 (continuous attribute transformation). The transformation results are effectively applied to the LR, MPL, and NB methods for datasets with continuous and categorical mixed attributes. [ABSTRACT FROM AUTHOR] |
|
Copyright of International Journal on Advanced Science, Engineering & Information Technology is the property of INSIGHT - Indonesian Society for Knowledge & Human Development and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.) |
| قاعدة البيانات: |
Complementary Index |
