Reduction of Response Variable Influential Outliers Using M-Estimation in the Next Day Prediction of Ground-Level Ozone Concentration
Keywords:Secondary pollutant, prediction, tuning constant, concentration
Ground-level ozone concentration (O3) is a second significant air pollutant in Malaysia after particulate matter 10 micrometres or less in diameter (PM10) concentration. It is a secondary pollutant that created by photochemical reaction of primary pollutant such as volatile organic compound (VOCs) and nitrogen oxides (NOx) under the influence of solar radiation (UVB). O3 photochemical reactions used solar radiation with certain wavelength as the catalyst. In statistical analysis of prediction, the concentration level of O3 contains the influential outliers due to several factors such as offense in data recording and sampling, the error in data acquisition or data management and the damage of monitoring instrument in data recording that can lead to misleading result or information. The objective of this study is to predict the level of O3 concentration for next day (D+1) by using predictors of wind speed (WS), temperature (T), relative humidity (RH), nitric oxide (NO), sulphur dioxide (SO2), nitrogen dioxide (NO2), ozone (O3) and carbon monoxide (CO) for selected urban area of Shah Alam by the method of minimizing influential outliers from response variable using M-estimation. The influential outliers from response variable is minimized using tuning constant approached at 95% level of efficiency. The improvement has been proved when Fair method has minimized 5.34% influential outliers from response variable and the average accuracy of the model is 0.5134
How to Cite
Open access licenses
Open Access is by licensing the content with a Creative Commons (CC) license.
This work is licensed under a Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International License.