Random Forest Implementation for Air Pollution Standard Index Classification in DKI Jakarta 2022

Authors

  • Hanifa Hasna Universitas Negeri Padang
  • Nonong Amalita
  • Dony Permana
  • Admi Salma

DOI:

https://doi.org/10.24036/ujsds/vol2-iss2/173

Abstract

Air pollution is a serious challenge in various cities, including DKI Jakarta. Based on measurements of the Air Pollution Standard Index carried out by the DKI Jakarta Environmental Service, the air quality in DKI Jakarta is considered moderate to unhealthy. Deteriorating air quality in the Jakarta metropolitan area is very dangerous for humans and living things. Therefore, to prevent the problem, the classification of air quality based on pollutant content is carried out using Random Forest (RF). The application of RF will form several trees that can provide better predictions and are able to produce low errors. The result of this study obtained optimal tree formation, namely tree formation using a combination of mtry (any input variables randomly selected in one sorting node)=2 and ntree (number of trees in the forest) as many as 5000 trees. The resulting accuracy was 99.17% with an OOB error rate of 0.83%. This research identifies that particulate pollutants are the main factor causing air pollution in DKI Jakarta. Based on these results, it shows that RF is able to provide accurate predictions about the level of air pollution in DKI Jakarta and can be identify important factors that affect air pollution.

Published

2024-05-31

How to Cite

Hasna, H., Nonong Amalita, Dony Permana, & Admi Salma. (2024). Random Forest Implementation for Air Pollution Standard Index Classification in DKI Jakarta 2022. UNP Journal of Statistics and Data Science, 2(2), 226–233. https://doi.org/10.24036/ujsds/vol2-iss2/173

Most read articles by the same author(s)

1 2 3 4 5 6 7 8 > >>