Price Forecasting Study by Machine Learning Using Weka Program
Article Sidebar
Main Article Content
Abstract
This research aimed to study machine learning models for price forecasting, using the Weka program (version 3.8). The objectives were to 1) create price forecasting models utilizing machine learning techniques and 2) create models that are effective in forecasting prices of fruits and vegetables, in order to support the systemic planning of future orders. Future price changes would be calculated by comparing the (Mean Absolute Error: MAE) values before and after adjusting parameters for 3 algorithms - Gaussian Distributor, Linear Regression and SMOreg technique, using the Weka program (version 3.8). After using the data of fruits and vegetables goes back for 6 years to create models to predict future prices, the results revealed that SMOreg algorithm gave the smallest total mean absolute error, before adjusting any parameters. Measuring performance with the mean absolute error (actual value versus forecast) gave a result of 2.206 and the Root Mean Square Error was 3.594. It was found that the model could be saved costs of 50,000 Baht per month.
Article Details
This work is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License.
Articles published in Journal of Industrial Technology Ubon Ratchathani Rajabhat University both hard copy and electronically are belonged to the Journal.
References
Thongkam P. Ensemble Algorithm for Feature Selection [Thesis]. Pathum thani; Thammasat University; 2016. (in Thai)
Sermpongpan C. Forcasting of cement consumption in thailand [Thesis]. Pathum thani; Thammasat University; 2017. (in Thai)
Bundasak S. Forecasting golf ball consumption with time series analysis [Thesis]. Chonburi; Kasetsart University; 2016. (in Thai)
Kapol W. Forecasting of power outages cause using data mining techniques a case study of provincial electricity [Thesis]. Pathum thani; Thammasat University; 2015. (in Thai)
Vilailuck S, Jaroenpuntaruk V, Wichadakul D. Utilizing Data Mining Techniques to Forecast Student Academic Achievement of Kasetsart University Laboratory School Kamphaeng Saen Campus Educational Research and Development Center. Veridian E-Journal, Science and Technology Silpakorn University. 2015; 2(2): 1-17.
Ngamsuk W. Time series forecasting by the comparision of classical and box-jenkinsmethods case study the number of accidents in the country. [Thesis]. Chonburi: Burapha University; 2012. (in Thai)
Konchalad P. The total energy consumption forecasting in Thailand by Box-Jenkins method and artificial neural network [Thesis]. Mahasarakham: Mahasarakham University; 2011. (in Thai)
Polasa N. To sell product modeling and sales forecasting of life insurance using data mining techniques Case study: Life insurance company [Thesis]. Pathum thani; Thammasat University; 2015. (in Thai)
Kaewhawong N. Forecasting Electricity Consumption of Thailand by Using SARIMA and Regression Models with ARMA Errors. Thai Journal of Science and Technology. 2015; 4(1): 24-36.
Vonganansup S, Yampaka T, Moosika O. Applied Data Mining Techniques To Analize Pre-Test And Skill-Test For Increased Student Potential. Rajamangala University of Technology Tawan-ok Research Journal. 2016; 5(1): 12-9.