TY - GEN
T1 - Spatial-Temporal pattern analysis and prediction of air quality in Taiwan
AU - Soh, Ping Wei
AU - Chen, Kai Hsiang
AU - Huang, Jen Wei
AU - Chu, Hone Jay
N1 - Funding Information:
The authors would like to thank the Ministry of Science and Technology, Taiwan, for financially supporting this research under Contract No. MOST 105-EPA-F-007-004. We are grateful to the Taiwan Environmental Protection Administration (TWEPA) for providing the monitoring data used in this study.
PY - 2017/10/18
Y1 - 2017/10/18
N2 - This study explores the spatial-temporal patterns of particulate matter (PM) in Taiwan. Probability map of PM and daily patterns are discussed in this study. Data mining provides more detailed spatial-temporal information for PM variations and trends. The proposed model will show that data mining provides a relatively high goodness of fit and sufficient space-time explanatory power, particularly air pollution frequency and affect areas. In the proposed model, a method using Dynamic Time Warping is proposed to analyse temporal similarity between stations. The proposed model can eliminate global effect on a single station through the performance of multiple stations. The proposed model will further be used for prediction of PM2.5. The prediction results will discuss the spatial-temporal relations between stations. This study will investigate the distribution of PM and its cyclicality.
AB - This study explores the spatial-temporal patterns of particulate matter (PM) in Taiwan. Probability map of PM and daily patterns are discussed in this study. Data mining provides more detailed spatial-temporal information for PM variations and trends. The proposed model will show that data mining provides a relatively high goodness of fit and sufficient space-time explanatory power, particularly air pollution frequency and affect areas. In the proposed model, a method using Dynamic Time Warping is proposed to analyse temporal similarity between stations. The proposed model can eliminate global effect on a single station through the performance of multiple stations. The proposed model will further be used for prediction of PM2.5. The prediction results will discuss the spatial-temporal relations between stations. This study will investigate the distribution of PM and its cyclicality.
UR - http://www.scopus.com/inward/record.url?scp=85039911764&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=85039911764&partnerID=8YFLogxK
U2 - 10.1109/UMEDIA.2017.8074094
DO - 10.1109/UMEDIA.2017.8074094
M3 - Conference contribution
AN - SCOPUS:85039911764
T3 - Ubi-Media 2017 - Proceedings of the 10th International Conference on Ubi-Media Computing and Workshops with the 4th International Workshop on Advanced E-Learning and the 1st International Workshop on Multimedia and IoT: Networks, Systems and Applications
BT - Ubi-Media 2017 - Proceedings of the 10th International Conference on Ubi-Media Computing and Workshops with the 4th International Workshop on Advanced E-Learning and the 1st International Workshop on Multimedia and IoT
PB - Institute of Electrical and Electronics Engineers Inc.
T2 - 10th International Conference on Ubi-Media Computing and Workshops, Ubi-Media 2017
Y2 - 1 August 2017 through 4 August 2017
ER -