Implementasi Synthetic Minority Oversampling Technique (SMOTE) Pada Algoritma Extreme Gradient Boosting (XGBOOST) Untuk Klasifikasi Indeks Standar Pencemaran Udara (ISPU)

Sajiwo, Achmad Fauzihan Bagus (2024) Implementasi Synthetic Minority Oversampling Technique (SMOTE) Pada Algoritma Extreme Gradient Boosting (XGBOOST) Untuk Klasifikasi Indeks Standar Pencemaran Udara (ISPU). Undergraduate thesis, UPN Veteran Jawa Timur.

	Text (Cover) 20081010069.-cover.pdf Download (414kB)
	Text (Bab 1) 20081010069.-bab1.pdf Download (312kB)
	Text (Bab 2) 20081010069.-bab2.pdf Restricted to Repository staff only until 24 July 2026. Download (521kB)
	Text (Bab 3) 20081010069.-bab3.pdf Restricted to Repository staff only until 24 July 2026. Download (457kB)
	Text (Bab 4) 20081010069.-bab4.pdf Restricted to Repository staff only until 24 July 2026. Download (948kB)
	Text (Bab 5) 20081010069.-bab5.pdf Download (171kB)
	Text (Daftar pustaka) 20081010069.-daftarpustaka.pdf Download (409kB)

Abstract

Air pollution is the entry of harmful substances into the atmosphere, which can be caused by human actions, whether intentional or unintentional, as well as by natural events. According to the Air Quality Live Index (AQLI) in April 2021, DKI Jakarta, as the nation's capital, is in sixth place in the world with the city with the worst air quality level. To deal with the problem of air pollution which continues to worsen, appropriate and effective action needs to be taken. One of them is conducting research on the classification of the air pollution standard index (ISPU). Implementing the ISPU classification requires a method that can process and analyze data patterns from sensors that measure air pollutant levels. The method used in this research is eXtreme Gradient Boosting (XGBoost). To help balance the data, this research used Synthetic Minority Over-sampling Technique (SMOTE). The data used is the DKI Jakarta ISPU for 2022-2023 which comes from the Satu Data Jakarta website: https://satudata.jakarta.go.id/home. The results of the standard air pollution index classification using the eXtreme Gradient Boosting algorithm with Synthetic Minority Over-sampling Technique, obtained an accuracy of 99.63%. From the confusion matrix calculations, precision, recall and f1-score values are obtained. For class 0, the precision value is 99%, the recall is 100% and the f1-score is 100%. For class 1, the precision score was 100%, the recall was 99% and the f1-score was 100%. For class 2, the precision score was 100%, the recall was 100% and the f1-score was 100%.

Item Type:

Thesis (Undergraduate)

Contributors: