Forecasting the air pollution index using artifical neural network at Muar, Johor, Malaysia / Ahmad Farid Rasdi
Clean and quality air is an essential element in maintaining a healthy quality of life. Air pollution is a serious issue that should be addressed by everyone around the world as it is one of the most important factors contributing to the quality of life and the environment. In addition, there is a s...
Saved in:
| Main Author: | |
|---|---|
| Format: | Thesis |
| Language: | English |
| Published: |
2021
|
| Subjects: | |
| Online Access: | http://ir.uitm.edu.my/id/eprint/44059/1/44059.pdf http://ir.uitm.edu.my/id/eprint/44059/ |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Summary: | Clean and quality air is an essential element in maintaining a healthy quality of life. Air pollution is a serious issue that should be addressed by everyone around the world as it is one of the most important factors contributing to the quality of life and the environment. In addition, there is a simple way to describe the air quality known as Air Pollution Index (API). With the index reference reading system, the API can easily detect changes in air quality. This study mainly focuses on forecasting the Air Pollution Index. In this study, secondary data was used which is obtained from the Department of Environment (DOE) regarding the Air Pollution Index in Malaysia. The dataset is the daily dataset Air Pollution Index (API) at Muar, Johor, Malaysia. The data is taken from the 1st of January 2015 to the 31st of December 2015. The method that was used in this study named Artificial Neural Network (ANN). Warren McCulloch and Walter Pitts developed this model by constructing a neural network computer model based on algorithms and mathematics and they are known as threshold logic. This study shows that ANN was conducted using the software named R Studio. It is shown that ANN was more accurately to be used as a forecasting method and to improve the accuracy of the forecasting compare to Naïve, Mean and ARIMA model using the lowest measures error which are Mean Error (ME), Root Mean Squared Error (RMSE), Mean Absolute Error (MAE), Mean Absolute Percentage Error (MAPE), and Mean Absolute Scaled Error (MASE). Besides, this study may also help the public to know the forecasted value API for the next three days. |
|---|