Sains Malaysiana 51(11)(2022): 3785-3793

http://doi.org/10.17576/jsm-2022-5111-22

Peramalan Kualiti Udara menggunakan Kaedah Pembelajaran Mendalam Rangkaian Perlingkaran Temporal (TCN)

(Air Quality Forecasting using Temporal Convolutional Network (TCN) Deep Learning Method)

MOHD AFTAR ABU BAKAR^*, NORATIQAH MOHD ARIFF, SAKHINAH ABU BAKAR, GOH PEI CHI & RAMYAH RAJENDRAN

Jabatan Sains Matematik, Fakulti Sains dan Teknologi, Universiti Kebangsaan Malaysia, 43600 UKM Bangi, Selangor Darul Ehsan, Malaysia

Received:13 March 2022/Accepted: 4 July 2022

Abstrak

Kajian ini bertujuan untuk membina model kualiti udara untuk meramalkan kepekatan bahan pencemar udara di Malaysia. Kaedah peramalan yang dipilih dalam kajian ini adalah suatu teknik pembelajaran mendalam iaitu Rangkaian Perlingkaran Temporal (TCN). Set data yang digunakan adalah siri masa zarahan terampai bersaiz diameter lebih kecil atau sama dengan 10 mikrometer (PM₁₀) yang diperoleh daripada Jabatan Alam Sekitar Malaysia dari 5 Julai 2017 hingga 31 Januari 2019. Data daripada lima stesen pemantauan kualiti udara di Semenanjung Malaysia dipilih untuk kajian ini. Bagi tujuan perbandingan, kaedah rangkaian memori jangka pendek panjang (LSTM) juga digunakan dalam kajian ini yang mana ketepatan antara kedua-dua model dibandingkan. Secara amnya, nilai model ramalan daripada kedua-dua model adalah menghampiri data asal. Walau bagaimanapun, model yang dibina dengan kaedah TCN adalah lebih baik berbanding model LSTM dari segi ketepatan nilai ramalan. Kajian ini menunjukkan bahawa TCN merupakan teknik yang sesuai digunakan dalam peramalan data siri masa bagi kualiti udara di Semenanjung Malaysia.

Kata kunci: Kualiti udara; pembelajaran mendalam; PM₁₀; Rangkaian Perlingkaran Temporal (TCN)

Abstract

This study aims to build an air quality model to predict pollutant concentrations in Malaysia. The method chosen in this study is one of the deep learning techniques which is the temporal convolution network (TCN). The data set used is particulate matter with diameter of 10 micrometers or less (PM₁₀) time series which is obtained from the Department of Environment Malaysia from 5^th July 2017 to 31^st January 2019. Data from five air quality monitoring stations in Peninsular Malaysia were selected for this study. The long-short term memory network (LSTM) is also used in this study for the purpose of accuracy comparison between the two models. Overall, the forecast values from both models are approximately close to the original data. However, the TCN model is better in terms of the forecast accuracy. This study shows that TCN is a suitable technique that can be used for forecasting air quality time series data in Peninsular Malaysia.

Keywords: Air quality; deep learning; PM₁₀; Temporal Convolutional Network (TCN)

REFERENCES

Awang, M.B., Jaafar, A.B., Abdullah, A.M., Ismail, M.B., Hassan, M.N., Abdullah, R., Johan, S. & Noor, H. 2000. Air quality in Malaysia: Impacts, management issues and future challenges. Respirology 5(2): 183-196.

Azid, A., Juahir, H., Ezani, E., Toriman, M.E., Endut, A., Rahman, M.N.A., Yunus, K., Kamarudin, M.K.A., Hasnam, C.N.C., Saudi, A.S.M. & Umar, R. 2015. Identification source of variation on regional impact of air quality pattern using chemometric. Aerosol and Air Quality Research 15(4): 1545-1558.

Bakar, M.A.A., Ariff, N.M., Nadzir, M.S.M., Wen, O.L. & Suris, F.N.A. 2022. Prediction of multivariate air quality time series data using long short-term memory network. Malaysian Journal of Fundamental and Applied Sciences 18: 52-59.

Carreras, M., Deriu, G., Raffo, L., Benini, L. & Meloni, P. 2020. Optimizing Temporal Convolutional Network inference on FPGA-based accelerators. IEEE Journal on Emerging and Selected Topics in Circuits and Systems 10(3): 348-361.

Chen, J., Chen, D. & Liu, G. 2020. Using temporal convolution network for remaining useful lifetime prediction. Engineering Reports 2020: e12305.

Chung, J., Gulcehre, C., Cho, K. & Bengio, Y. 2014. Empirical evaluation of gated recurrent neural networks on sequence modeling. arXiv preprint arXiv: 1412.3555.

Dockery, D.W., Pope, C.A., Xu, X., Spengler, J.D., Ware, J.H., Fay, M.E., Ferris Jr., B.G. & Speizer, F.E. 1993. An association between air pollution and mortality in six US cities. New England Journal of Medicine 329(24): 1753-1759.

Elman, J.L. 1990. Finding structure in time. Cognitive Science 14(2): 179-211.

Freeman, B.S., Taylor, G., Gharabaghi, B. & Thé, J. 2018. Forecasting air quality time series using deep learning. Journal of the Air & Waste Management Association 68(8): 866-886.

Halim, N.D.A., Latif, M.T., Mohamed, A.F., Maulud, K.N.A., Idrus, S., Azhari, A., Othman, M. & Sofwan, N.M. 2020. Spatial assessment of land use impact on air quality in mega urban regions, Malaysia. Sustainable Cities and Society 63: 102436.

He, Y. & Zhao, J. 2019. Temporal convolutional networks for anomaly detection in time series. Journal of Physics: Conference Series 1213(4): 042050.

Kim, H.Y. & Kim, D. 2020. Prediction of mutation effects using a deep temporal convolutional network. Bioinformatics 36(7): 2047-2052.

Khairi, S.S.M., Bakar, M.A.A., Alias, M.A., Bakar, S.A., Liong, C.Y., Rosli, N. & Farid, M. 2021. Deep learning on histopathology images for breast cancer classification: A bibliometric analysis. Healthcare 10(10): 1-22.

Lai, G., Chang, W.C., Yang, Y. & Liu, H. 2018. Modeling long-and short-term temporal patterns with deep neural networks. In The 41st International ACM SIGIR Conference on Research & Development in Information Retrieval. pp. 95-104.

Latif, M.T., Othman, M., Idris, N., Juneng, L., Abdullah, A.M., Hamzah, W.P., Khan, M.F., Sulaiman, N.M.N., Jewaratnam, J., Aghamohammadi, N. & Sahani, M. 2018. Impact of regional haze towards air quality in Malaysia: A review. Atmospheric Environment 177: 28-44.

Li, B. & Sim, K.C. 2014. A spectral masking approach to noise-robust speech recognition using deep neural networks. IEEE/ACM Transactions on Audio, Speech, and Language Processing 22(8): 1296-1305.

Nadzir, M.S.M., Nor, M.Z.M., Nor, M.F.F.M., Wahab, M.I.A., Ali, S.H.M., Otuyo, M.K., Bakar, M.A.A., Saw, L.H., Majumdar, S., Ooi, M.C.G., Mohamed, F., Hisham, B.A., Hamid, H.H.A., Khaslan, Z., Ariff, N.M., Anuar, J., Tok, G.R., Ya’akop, N.A. & Meswn, M.M. 2021. Risk assessment and air quality study during different phases of COVID-19 lockdown in an urban area of Klang Valley, Malaysia. Sustainability 13: 12217.

Nadzir, M.S.M., Ooi, M.C.G., Alhasa, K.M., Bakar, M.A.A., Mohtar, A.A.A., Nor, M.F.F.M., Latif, M.T., Hamid, H.H.A., Ali, S.H.M., Ariff, N.M., Anuar, J., Ahmad, F., Azhari, A., Hanif, N.M., Subhi, M.A., Othman, M. & Nor, M.Z.M. 2020. The impact of Movement Control Order (MCO) during pandemic COVID-19 on local air quality in an urban area of Klang Valley, Malaysia. Aerosol and Air Quality Research 20: 1237-1248.

Purwins, H., Li, B., Virtanen, T., Schlüter, J., Chang, S.Y. & Sainath, T. 2019. Deep learning for audio signal processing. IEEE Journal of Selected Topics in Signal Processing 13(2): 206-219.

Sethi, J.K. & Mittal, M. 2020. Analysis of air quality using univariate and multivariate time series models. In 2020 10th International Conference on Cloud Computing, Data Science & Engineering (Confluence) IEEE. pp. 823-827.

Sun, J., Luo, X., Gao, H., Wang, W., Gao, Y. & Yang, X. 2020. Categorizing malware via a word2vec-based temporal convolutional network scheme. Journal of Cloud Computing 9(1): 1-14.

Suris, F.N.A., Bakar, M.A.A., Ariff, N.M., Mohd Nadzir, M.S. & Ibrahim, K. 2022. Malaysia PM₁₀ air quality time series clustering based on dynamic time warping. Atmosphere 13: 503.

Yan, J., Mu, L., Wang, L., Ranjan, R. & Zomaya, A.Y. 2020. Temporal convolutional     networks for the advance prediction of ENSO. Scientific Reports 10(1): 1-15.

Zhao, J., Deng, F., Cai, Y. & Chen, J. 2019. Long short-term memory - Fully connected (LSTM-FC) neural network for PM2.5 concentration prediction. Chemosphere 220: 486-492.

^*Corresponding author; email: aftar@ukm.edu.my