Abstract

Forecasting is an instrument of decision-making that makes predictions or estimates about the future based on historical data. Identifying a suitable strategy for forecasting a time series amongst the classical techniques (e.g., exponential smoothing, Auto-Regressive Integrated Moving Average (ARIMA)), Neural approach, and Support Vector Regression (SVR) - another widely used and popular machine learning-based approach, is challenging. The present work aimed at providing a simple (implementation wise), efficient (forecast accuracy wise), and state-of-art Multi-Layer Perceptron (MLP) approach for some selected macroeconomic indices (Wholesale Price Index - i.e., WPI) in India. We looked at the WPIs with non-linear trends identified using the curve-fit method. It's known that the diverse Indian chemical industry contributes notably to India's economic development. In this work, we analyzed the WPI of seventy-seven commodities/items of the "manufacture of chemicals and chemical products" group in India. We detected the indices having non-linear trends by applying the curve-fit method. The curve-fit approach based on statistical rigor identifies the non-linear WPIs. Twenty-five out of seventy-seven indices exhibits non-linear trends. We developed a forecasting approach employing the MLP for these twenty-five non-linear WPIs. The proposed-MLP optimized by hyperparameter tuning offers high accuracy, prediction reliability, and prediction acceptability for all non-linear WPIs. The forecasting performances of the proposed-MLP compared with regression models (Linear, Quadratic, Cubic, Logarithmic, Exponential), exponential smoothing (Holt linear trend, Holt exponential trend, Holt-Winters), state-of-art Auto-ARIMA, and SVR. The MLP outperformed them all. In terms of Mean Absolute Percentage Error (MAPE), the MLP outperform Linear in 88%, Quadratic in 92%, Cubic in 88%, Logarithmic in 72%, exponential in 88%, Holt Linear in 80%, Holt Exponential in 76%, Holt-Winters in 72%, Auto-ARIMA in 56%, and SVR in 56% of cases. We suggest the application of the proposed approach as an alternative for forecasting these twenty-five non-linear WPIs.

Keyword

Curve fitting, Multilayer perceptron, Wholesale price index, ARIMA, Exponential smoothing, Support vector regression.

Cite this article

Das D, Chakrabarti S

Refference

[1][1]Wadi SA, Almasarweh M, Alsaraireh AA, Aqaba J. Predicting closed price time series data using ARIMA model. Modern Applied Science. 2018; 12(11):181-5.

[2][2]Almasarweh M, Alwadi S. ARIMA model in predicting banking stock market data. Modern Applied Science. 2018; 12(11):309-12.

[3][3]Miller JW. ARIMA time series models for full truckload transportation prices. Forecasting. 2019; 1(1):121-34.

[4][4]Talwar A, Goyal CK. A comparative study of various exponential smoothing models for forecasting coriander price in Indian commodity market. International Bulletin of Management and Economics. 2019:143-55.

[5][5]Suppalakpanya K, Nikhom R, Booranawong T, Booranawong A. Study of several exponential smoothing methods for forecasting crude palm oil productions in Thailand. Current Applied Science and Technology. 2019; 19(2):123-39.

[6][6]Şahinli MA. Potato price forecasting with holt-winters and ARIMA methods: a case study. American Journal of Potato Research. 2020; 97(4):336-46.

[7][7]Claveria O, Monte E, Torra S. Modelling tourism demand to Spain with machine learning techniques. The impact of forecast horizon on model selection. arXiv preprint arXiv:1805.00878. 2018.

[8][8]He XJ. Crude oil prices forecasting: time series vs. SVR models. Journal of International Technology and Information Management. 2018; 27(2):25-42.

[9][9]Kuizinienė D, Mackutė-varoneckienė A, Krilavičius T. Cryptocurrencies short-term forecast: application of ARIMA, GARCH and SVR models. In international conference on information technologies, Kaunas, Lithuania. 2019 (pp.70-3).

[10][10]Abidoye RB, Chan AP, Abidoye FA, Oshodi OS. Predicting property price index using artificial intelligence techniques: evidence from Hong Kong. International Journal of Housing Markets and Analysis. 2019; 12(6):1072-92.

[11][11]Airlangga G, Rachmat A, Lapihu D. Comparison of exponential smoothing and neural network method to forecast rice production in Indonesia. Telkomnika. 2019; 17(3):1367-75.

[12][12]Spiliotis E, Doukas H, Assimakopoulos V, Petropoulos F. Forecasting week-ahead hourly electricity prices in belgium with statistical and machine learning methods. Mathematical Modelling of Contemporary Electricity Markets 2021:59-74. Academic Press.

[13][13]Hossain MF, Nandi DC, Uddin KM. Prediction of banking sectors financial data of Dhaka stock exchange using autoregressive integrated moving average approach. International Journal of Material and Mathematical Science. 2020; 2(4):64-70.

[14][14]Rohmah MF, Putra IK, Hartati RS, Ardiantoro L. Comparison four kernels of SVR to predict consumer price index. In journal of physics: conference series 2021 (pp. 1-9). IOP Publishing.

[15][15]Marini F. 3.14 – Neural networks. Comprehensive Chemometrics. 2009 : 477-505. Elsevier.

[16][16]Bertolaccini L, Solli P, Pardolesi A, Pasini A. An overview of the use of artificial neural networks in lung cancer research. Journal of Thoracic Disease. 2017; 9(4):924-31.

[17][17]https://towardsdatascience.com/introduction-to-neural-networks-advantages-and-applications-96851bd1a207. Accessed 10 April 2021.

[18][18]Azari A. Bitcoin price prediction: an ARIMA approach. arXiv preprint arXiv:1904.05315. 2019.

[19][19]Zhu W, Li X, Sun B. Research and prediction on chinas novel coronavirus (2019-nCoV/COVID-19) epidemic-based on time series ARIMA model. In international conference on education, management, computer and society 2020 (pp.310-16).

[20][20]Katoch R, Sidhu A. An application of ARIMA model to forecast the dynamics of COVID-19 epidemic in India. Global Business Review. 2021:1-14.

[21][21]Rasheed A, Ullah MA, Uddin I. PKR exchange rate forecasting through univariate and multivariate time series techniques. NICE Research Journal. 2020; 13(4):49-67.

[22][22]Sokannit P, Chujai P. Forecasting household electricity consumption using time series models. International Journal of Machine Learning and Computing. 2021; 11(6):380-6.

[23][23]Li R, Li X, Lu Z. Forecast of gross output value of agriculture and forestry in guangxi based on holt-winters model. In journal of physics: conference series 2021 (pp.1-7). IOP Publishing.

[24][24]Ali K. Forecasting Analysis of share price index in construction companies registered in indonesia stock exchange 2015-2019. Journal of Economics Research and Social Sciences. 2021; 5(1):42-63.

[25][25]Carrasco R, Astudillo G, Soto I, Chacon M, Fuentealba D. Forecast of copper price series using vector support machines. In international conference on industrial technology and management 2018 (pp. 380-4). IEEE.

[26][26]Alegado RT, Tumibay GM. Statistical and machine learning methods for vaccine demand forecasting: a comparative analysis. Journal of Computer and Communications. 2020; 8(10):37.

[27][27]Talkhi N, Fatemi NA, Ataei Z, Nooghabi MJ. Modeling and forecasting number of confirmed and death caused COVID-19 in IRAN: a comparison of time series forecasting methods. Biomedical Signal Processing and Control. 2021:1-8.

[28][28]https://data.gov.in/resources/wholesale-price-index-base-year-2011-12-upto-may-2017. Accessed 01 March 2021.

[29][29]Team RC. R: a language and environment for statistical computing. 2013.

[30][30]https://cran.r-project.org/web/ packages/ nnfor/ index.html. Accessed 10 April 2021.

[31][31]Das D. Data mining of indian stock market from April, 2015 to March, 2016 using curve fitting technique. International Research Journal of Engineering and Technology. 2016; 3(5):2564-70.

[32][32]Juarna A. One year stock price prediction and its validity using least square method in MATLAB. International Journal of Advanced Research. 2017; 5(2): 1641-48.

[33][33]Awan TM, Aslam F. Prediction of daily COVID-19 cases in European countries using automatic ARIMA model. Journal of Public Health Research. 2020; 9(3):227-33.

[34][34]Yermal L, Balasubramanian P. Application of auto arima model for forecasting returns on minute wise amalgamated data in NSE. In international conference on computational intelligence and computing research 2017 (pp. 1-5). IEEE.

[35][35]https://pkg.robjhyndman.com/forecast/. Accessed 10 April 2021.

[36][36]Hyndman RJ, Khandakar Y. Automatic time series forecasting: the forecast package for R. Journal of Statistical Software. 2008; 27(1):1-22.

[37][37]David M, Evgenia D, Kurt H, Andreas W, Friedrich L. E1071: misc functions of the department of statistics, probability theory group (Formerly: E1071), TU Wien. R Package Version. 2019.

[38][38]https://www.rdocumentation.org/packages/DescTools/versions/0.99.36/topics/Measures%20of%20Accuracy. Accessed 02 April 2021.

[39][39]Oshodi OS, Ejohwomu OA, Famakin IO, Cortez P. Comparing univariate techniques for tender price index forecasting: box-jenkins and neural network model. Construction Economics and Building. 2017; 17(3):109-23.

[40][40]Yadav V, Nath S. Novel hybrid model for daily prediction of PM 10 using principal component analysis and artificial neural network. International Journal of Environmental Science and Technology. 2019; 16(6):2839-48.

[41][41]Fan RY, Ng ST, Wong JM. Reliability of the Box–Jenkins model for forecasting construction demand covering times of economic austerity. Construction Management and Economics. 2010; 28(3):241-54.

[42][42]Khashei M, Hajirahimi Z. A comparative study of series ARIMA/MLP hybrid models for stock price forecasting. Communications in Statistics-Simulation and Computation. 2019; 48(9):2625-40.

[43][43]Lu W, Li J, Li Y, Sun A, Wang J. A CNN-LSTM-based model to forecast stock prices. Complexity. 2020.

[44][44]Herrera GP, Constantino M, Tabak BM, Pistori H, Su JJ, Naranpanawa A. Long-term forecast of energy commodities price using machine learning. Energy. 2019; 179:214-21.