Asst. Prof. Nantachai Kantanantha, Ph.D.
- 5th Floor of Engineering 4 Bldg., Room 508
- +66-2218-6820
- nantachai.k@chula.ac.th
Education
Ph.D. in Industrial Engineering
Georgia Institute of Technology, United States, 2007
M.S. in Industrial Engineering
Georgia Institute of Technology, United States, 2001
B.Eng. in Industrial Engineering
Chulalongkorn University, Thailand
Expertise
Statistics & Data Analysis
Publications
2016
Onuma Kosanan, Nantachai Kantanantha
vol. 8-10 March 2016, IEOM Society, 2016, (Cited by: 0).
@conference{Kosanan20161167,
title = {A hybrid particle swarm optimization algorithm and support vector machine model for agricultural statistic of Thailand forecasting},
author = {Onuma Kosanan and Nantachai Kantanantha},
url = {https://www.scopus.com/inward/record.uri?eid=2-s2.0-85018390169&partnerID=40&md5=9ca9a5448201ce08da2270ea9b4187b6},
year = {2016},
date = {2016-01-01},
journal = {Proceedings of the International Conference on Industrial Engineering and Operations Management},
volume = {8-10 March 2016},
pages = {1167 – 1178},
publisher = {IEOM Society},
abstract = {The objective of this research is to construct a Thailand's Para rubber production forecasting model. It will be advantageous to farmers, entrepreneurs and other organizations for the right planning and decision making in order to prepare themselves to be ready for the modernized global economics trends which will affect to Thailand's agricultural economy. Four forecasting techniques used in this research artificial neural network (ANN), particle swarm optimization algorithm (PSO), support vector machine (SVM) and hybrid model PSO and SVM. The mean absolute percentage error is used to identify the most appropriate model. The results of the research show that the hybrid PSO&SVM model obtains the lowest mean absolute percentage error of 0.0040%, while the particle swarm optimization model, support vector machine model and artificial neural network model have mean absolute percentage error of 0.0388%, 0.0388% and 0.0414% respectively. © IEOM Society International. © IEOM Society International.},
note = {Cited by: 0},
keywords = {},
pubstate = {published},
tppubtype = {conference}
}
The objective of this research is to construct a Thailand’s Para rubber production forecasting model. It will be advantageous to farmers, entrepreneurs and other organizations for the right planning and decision making in order to prepare themselves to be ready for the modernized global economics trends which will affect to Thailand’s agricultural economy. Four forecasting techniques used in this research artificial neural network (ANN), particle swarm optimization algorithm (PSO), support vector machine (SVM) and hybrid model PSO and SVM. The mean absolute percentage error is used to identify the most appropriate model. The results of the research show that the hybrid PSO&SVM model obtains the lowest mean absolute percentage error of 0.0040%, while the particle swarm optimization model, support vector machine model and artificial neural network model have mean absolute percentage error of 0.0388%, 0.0388% and 0.0414% respectively. © IEOM Society International. © IEOM Society International.
2014
Onuma Kosanan, Nantachai Kantanantha
Thailand’s Para rubber production forecasting comparison Conference
vol. 2210, no. January, Newswood Limited, 2014, (Cited by: 5).
@conference{Kosanan2014,
title = {Thailand's Para rubber production forecasting comparison},
author = {Onuma Kosanan and Nantachai Kantanantha},
editor = {Castillo O. and Ao S.I. and Lee J.-A. and Douglas C. and Feng D.D.},
url = {https://www.scopus.com/inward/record.uri?eid=2-s2.0-84938360201&partnerID=40&md5=162c634d135358a3e61e0b67febd8769},
year = {2014},
date = {2014-01-01},
journal = {Lecture Notes in Engineering and Computer Science},
volume = {2210},
number = {January},
publisher = {Newswood Limited},
abstract = {The objective of this research is to construct a Thailand's Para rubber production forecasting model. Three forecasting techniques used in this research are auto regressive integrated moving average (ARIMA), artificial neural network (ANN) and support vector machine (SVM). The mean absolute percentage error is used to identify the most appropriate model. The results of the research show that the artificial neural network model obtains the lowest mean absolute percentage error of 0.0037%, while the auto regressive integrated moving average and support vector machine have mean absolute percentage error of 0.0419% and 0.0434%, respectively.},
note = {Cited by: 5},
keywords = {},
pubstate = {published},
tppubtype = {conference}
}
The objective of this research is to construct a Thailand’s Para rubber production forecasting model. Three forecasting techniques used in this research are auto regressive integrated moving average (ARIMA), artificial neural network (ANN) and support vector machine (SVM). The mean absolute percentage error is used to identify the most appropriate model. The results of the research show that the artificial neural network model obtains the lowest mean absolute percentage error of 0.0037%, while the auto regressive integrated moving average and support vector machine have mean absolute percentage error of 0.0419% and 0.0434%, respectively.
Panudet Saengseeedam, Nantachai Kantanantha
Spatial time series forecasts based on Bayesian linear mixed models for rice yields in Thailand Conference
vol. 2210, no. January, Newswood Limited, 2014, (Cited by: 0).
@conference{Saengseeedam2014,
title = {Spatial time series forecasts based on Bayesian linear mixed models for rice yields in Thailand},
author = {Panudet Saengseeedam and Nantachai Kantanantha},
editor = {Castillo O. and Ao S.I. and Lee J.-A. and Douglas C. and Feng D.D.},
url = {https://www.scopus.com/inward/record.uri?eid=2-s2.0-84938335342&partnerID=40&md5=7c1e2b05d28cbba22d09dbd18777dbf0},
year = {2014},
date = {2014-01-01},
journal = {Lecture Notes in Engineering and Computer Science},
volume = {2210},
number = {January},
publisher = {Newswood Limited},
abstract = {Spatial time series forecasts using linear mixed models (LMMs) with spatial effects under a Bayesian framework are considered. The random effects are assumed to be normally distributed and the spatial effects are assumed to be CAR models. The proposed model is applied to the rice yields data in 19 Northeastern provinces in Thailand. It has a better performance, using the MAE criteria, compared to the existing simple exponential smoothing (ES) and autoregressive integrated moving average (ARIMA) models.},
note = {Cited by: 0},
keywords = {},
pubstate = {published},
tppubtype = {conference}
}
Spatial time series forecasts using linear mixed models (LMMs) with spatial effects under a Bayesian framework are considered. The random effects are assumed to be normally distributed and the spatial effects are assumed to be CAR models. The proposed model is applied to the rice yields data in 19 Northeastern provinces in Thailand. It has a better performance, using the MAE criteria, compared to the existing simple exponential smoothing (ES) and autoregressive integrated moving average (ARIMA) models.
2013
Pitsanu Tongkhow, Nantachai Kantanantha
Bayesian models for time series with covariates, trend, seasonality, autoregression and outliers Journal Article
In: Journal of Computer Science, vol. 9, no. 3, pp. 291 – 298, 2013, (Cited by: 3; All Open Access, Hybrid Gold Open Access).
@article{Tongkhow2013291,
title = {Bayesian models for time series with covariates, trend, seasonality, autoregression and outliers},
author = {Pitsanu Tongkhow and Nantachai Kantanantha},
url = {https://www.scopus.com/inward/record.uri?eid=2-s2.0-84880109258&doi=10.3844%2fjcssp.2013.291.298&partnerID=40&md5=2612917092dd5c795889a0018f9ea66a},
doi = {10.3844/jcssp.2013.291.298},
year = {2013},
date = {2013-01-01},
journal = {Journal of Computer Science},
volume = {9},
number = {3},
pages = {291 – 298},
abstract = {Bayesian methods furnish an attractive approach to time series data analysis. This article proposes the forecasting models that can detect trend, seasonality, auto regression and outliers in time series data related to some covariates. Cumulative Weibull distribution functions for trend, dummy variables for seasonality, binary selections for outliers and latent autoregression for autocorrelated time series data are used for the data analysis. The Gibbs sampling, a Markov Chain Monte Carlo (MCMC) algorithm, is used for the parameter estimation. The proposed models are applied to vegetable price time series data in Thailand. According to the RMSE, MAPE and MAE criteria for model comparisons, the proposed models provide the best results compared to the exponential smoothing models, SARIMA models and the Bayesian models with trend, auto regression and outliers. © 2013 Science Publications.},
note = {Cited by: 3; All Open Access, Hybrid Gold Open Access},
keywords = {},
pubstate = {published},
tppubtype = {article}
}
Bayesian methods furnish an attractive approach to time series data analysis. This article proposes the forecasting models that can detect trend, seasonality, auto regression and outliers in time series data related to some covariates. Cumulative Weibull distribution functions for trend, dummy variables for seasonality, binary selections for outliers and latent autoregression for autocorrelated time series data are used for the data analysis. The Gibbs sampling, a Markov Chain Monte Carlo (MCMC) algorithm, is used for the parameter estimation. The proposed models are applied to vegetable price time series data in Thailand. According to the RMSE, MAPE and MAE criteria for model comparisons, the proposed models provide the best results compared to the exponential smoothing models, SARIMA models and the Bayesian models with trend, auto regression and outliers. © 2013 Science Publications.
2012
Nantachai Kantanantha, Pornprasert Leelakriangsak
vol. 1, Newswood Limited, 2012, (Cited by: 0).
@conference{Kantanantha2012566,
title = {Comparison of cost estimates of electrical and communication system for industrial factory construction},
author = {Nantachai Kantanantha and Pornprasert Leelakriangsak},
editor = {Burgstone J. and Ao S.I. and Douglas C. and Grundfest W.S.},
url = {https://www.scopus.com/inward/record.uri?eid=2-s2.0-85049880275&partnerID=40&md5=86b445e30559bbb19640888e05873150},
year = {2012},
date = {2012-01-01},
journal = {Lecture Notes in Engineering and Computer Science},
volume = {1},
pages = {566 – 570},
publisher = {Newswood Limited},
abstract = {Construction cost estimation is essential for turnkey construction. The good estimate should be a fair price for both customer and construction company. This research aims to compare the cost estimates of electrical and communication system for industrial factory construction. Three forecasting methods compared in this research are Multiple Regression Analysis (MRA), Multiple Regression Analysis incorporating Genetic Algorithm (MRA-GA), and Neural Network (NN). The data sets are collected from 31 industrial factory projects constructed in Thailand between year 2005 and 2011 which are divided into 25 training data sets and 6 testing data sets. The selected input variables are area, cost percentage from copper, cost percentage from main equipment, cost percentage from labor, and air condition system. The results show that MRA-GA model provides slightly lower root mean squared error (RMSE) than MRA and NN models. © 2012 Newswood Limited. All rights reserved.},
note = {Cited by: 0},
keywords = {},
pubstate = {published},
tppubtype = {conference}
}
Construction cost estimation is essential for turnkey construction. The good estimate should be a fair price for both customer and construction company. This research aims to compare the cost estimates of electrical and communication system for industrial factory construction. Three forecasting methods compared in this research are Multiple Regression Analysis (MRA), Multiple Regression Analysis incorporating Genetic Algorithm (MRA-GA), and Neural Network (NN). The data sets are collected from 31 industrial factory projects constructed in Thailand between year 2005 and 2011 which are divided into 25 training data sets and 6 testing data sets. The selected input variables are area, cost percentage from copper, cost percentage from main equipment, cost percentage from labor, and air condition system. The results show that MRA-GA model provides slightly lower root mean squared error (RMSE) than MRA and NN models. © 2012 Newswood Limited. All rights reserved.
Pitsanu Tongkhow, Nantachai Kantanantha
Bayesian model for time series with trend, autoregression and outliers Conference
2012, (Cited by: 1).
@conference{Tongkhow201290,
title = {Bayesian model for time series with trend, autoregression and outliers},
author = {Pitsanu Tongkhow and Nantachai Kantanantha},
url = {https://www.scopus.com/inward/record.uri?eid=2-s2.0-84873400989&doi=10.1109%2fICTKE.2012.6408577&partnerID=40&md5=554097109f32168508927ea0b7eb4d5a},
doi = {10.1109/ICTKE.2012.6408577},
year = {2012},
date = {2012-01-01},
journal = {International Conference on ICT and Knowledge Engineering},
pages = {90 – 94},
abstract = {We propose the Bayesian forecasting model that can detect trend, autoregression, and outliers in the time series data. We use cumulative Weibull distribution function for trend, binary selection for outliers, and autoregression for related time series data. Gibbs sampling algorithm which is one of MCMC methods is used for parameter estimation. The proposed models are applied to the vegetable price time series data in Thailand. According to the RMSE, MAPE, and MAE criteria for the model comparison, the proposed model provides the best results compared to the exponential smoothing and SARIMA models. © 2012 IEEE.},
note = {Cited by: 1},
keywords = {},
pubstate = {published},
tppubtype = {conference}
}
We propose the Bayesian forecasting model that can detect trend, autoregression, and outliers in the time series data. We use cumulative Weibull distribution function for trend, binary selection for outliers, and autoregression for related time series data. Gibbs sampling algorithm which is one of MCMC methods is used for parameter estimation. The proposed models are applied to the vegetable price time series data in Thailand. According to the RMSE, MAPE, and MAE criteria for the model comparison, the proposed model provides the best results compared to the exponential smoothing and SARIMA models. © 2012 IEEE.
2010
Nantachai Kantanantha, Nicoleta Serban, Paul Griffin
Yield and price forecasting for stochastic crop decision planning Journal Article
In: Journal of Agricultural, Biological, and Environmental Statistics, vol. 15, no. 3, pp. 362 – 380, 2010, (Cited by: 44).
@article{Kantanantha2010362,
title = {Yield and price forecasting for stochastic crop decision planning},
author = {Nantachai Kantanantha and Nicoleta Serban and Paul Griffin},
url = {https://www.scopus.com/inward/record.uri?eid=2-s2.0-84864829028&doi=10.1007%2fs13253-010-0025-7&partnerID=40&md5=34f8a1fe0f4a6164776e17387895b6bf},
doi = {10.1007/s13253-010-0025-7},
year = {2010},
date = {2010-01-01},
journal = {Journal of Agricultural, Biological, and Environmental Statistics},
volume = {15},
number = {3},
pages = {362 – 380},
abstract = {The primary objective of this paper is to develop yield and price forecasting models employed in informed crop decision planning-a key aspect of effective farm management. For yearly yield prediction, we introduce a weather-based regression model with time-dependent varying coefficients. In order to allow for within-year climate variations, we predict yearly crop yield using weekly temperature and rainfall summaries resulting in a large number of correlated predictors. To overcome this difficulty, we reduce the space of predictors to a small number of uncorrelated predictors using Functional Principal Component Analysis (FPCA). For detailed price forecasting, we develop a futures-based model for long-range cash price prediction. In this model, the cash price is predicted as a sum of the nearby settlement futures price and the predicted commodity basis. We predict the one-year commodity basis as a mixture of historical basis data using a functional model-based approach. In both forecasting models, we estimate approximate prediction confidence intervals that are further integrated in a decision planning model. We applied our methods to corn yield and price forecasting for Hancock County in Illinois. Our forecasting results are more accurate in comparison to predictions based on existing methods. The methods introduced in this paper generally apply to other locations in the US and other crop types. The supplemental materials for this article are available online. © 2010 International Biometric Society.},
note = {Cited by: 44},
keywords = {},
pubstate = {published},
tppubtype = {article}
}
The primary objective of this paper is to develop yield and price forecasting models employed in informed crop decision planning-a key aspect of effective farm management. For yearly yield prediction, we introduce a weather-based regression model with time-dependent varying coefficients. In order to allow for within-year climate variations, we predict yearly crop yield using weekly temperature and rainfall summaries resulting in a large number of correlated predictors. To overcome this difficulty, we reduce the space of predictors to a small number of uncorrelated predictors using Functional Principal Component Analysis (FPCA). For detailed price forecasting, we develop a futures-based model for long-range cash price prediction. In this model, the cash price is predicted as a sum of the nearby settlement futures price and the predicted commodity basis. We predict the one-year commodity basis as a mixture of historical basis data using a functional model-based approach. In both forecasting models, we estimate approximate prediction confidence intervals that are further integrated in a decision planning model. We applied our methods to corn yield and price forecasting for Hancock County in Illinois. Our forecasting results are more accurate in comparison to predictions based on existing methods. The methods introduced in this paper generally apply to other locations in the US and other crop types. The supplemental materials for this article are available online. © 2010 International Biometric Society.