Open AccessPrediction algorithms to forecast air pollution in Delhi India on a decade
Author(s) -
Muhammad Rifki Taufik,
Eka Rosanti,
Tofan Agung Eka Prasetya,
Tri Wijayanti Septiarini
Publication year - 2020
Publication title -
journal of physics. conference series
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.21
H-Index - 85
eISSN - 1742-6596
pISSN - 1742-6588
DOI - 10.1088/1742-6596/1511/1/012052
Subject(s) - exponential smoothing , autoregressive integrated moving average , air pollution , air quality index , data set , time series , meteorology , computer science , statistics , algorithm , econometrics , machine learning , mathematics , geography , chemistry , organic chemistry
According to the WHO Global Ambient Air Quality Database in the past two years, there are more than 4300 cities and settlements in 108 countries where have nearly doubled, especially in Delhi, India. Preventing unwanted events is a mandatory crucial step by forecasting air pollution identifying air quality levels and recognizing the associated health impacts. Aim of this paper to forecast air pollution using four prediction models i.e. Naïve Bayesian, Auto Regressive Integrated Moving Average (ARIMA), Exponential Smoothing, and TBATS. The data were obtained from the official website of the Indian government where this research analyzed time-series data from 2005-2015 consisted of PM10, SO 2 , and NO 2 with time variables day, month, and year. The time series set was managed to be monthly in ten years. Moreover, the series was split into a training set and testing set with a ratio 75:25. The training set was utilized to build prediction models and the testing set would evaluate forecasting results. Forecasting results showed all models gave acceptable prediction and according to the error, the ARIMA and exponential smoothing models were the potential prediction model for air pollution data.