Search

Thailand's economy, like many other ASEAN countries, heavily relies on remittances from Thai overseas workers. Predicting the number of such workers in the coming years is crucial for stakeholders engaged in economic planning and growth. This study aimed to find a reliable statistical model for this purpose, utilizing data from pre-pandemic years to compare two predictive methods. The dataset spans 120 months, from January 2010 to December 2019, and was obtained from the Library System, Department of Employment of the Royal Thai Government. The chosen forecasting techniques are the Box-Jenkins method and Holt-Winter's Exponential Smoothing method. Primary parameters were first estimated based on the given data, and then evaluated through the Bayesian Information Criterion (BIC). To assess model performance, the dataset was split into two subsets. The primary parameters were applied to the first 60 months' data (January 2010 to December 2014) to estimate secondary parameters. The remaining 60 months' data were then used to evaluate how well each model predicts the response variable, measured by the prediction mean absolute error (PMAE) and prediction root mean squared error (PRMSE). Based on the findings, the Box-Jenkins method outperformed Holt-Winter's method in forecasting the monthly number of fresh Thai overseas workers. This study's insights can serve as a valuable template for predicting overseas workers' numbers in other ASEAN countries with socio-economic-cultural contexts similar to Thailand. Accurate predictions can aid decision-making and planning for sustainable economic development in the region. [ABSTRACT FROM AUTHOR]

This study applies statistical methods and deep learning techniques to forecast the daily average PM2.5 concentration in northern Thailand, where the concentration is usually high and exceeds the safe level. The data used in the analysis are collected from January 2018 to December 2020 from 16 air monitoring stations. The statistical methods used are Holt-Winters exponential smoothing (ETS), autoregressive integrated moving average (ARIMA), and dynamic linear model (DLM). The deep learning techniques considered in this study are the recurrent neural network (RNN) and long-short term memory (LSTM). To compare the predictive performance of both methods, we use the root mean square error (RMSE). The result indicates that statistical methods, especially ARIMA, perform better than the deep learning techniques in most stations. Moreover, LSTM tends to provide higher accuracy than the RNN, especially with more number of nodes. [ABSTRACT FROM AUTHOR]

Thailand is one of the countries where foot and mouth disease outbreaks have resulted in considerable economic losses. Forecasting is an important warning technique that can allow authorities to establish an FMD surveillance and control program. This study aimed to model and forecast the monthly number of FMD outbreak episodes (n-FMD episodes) in Thailand using the time-series methods, including seasonal autoregressive integrated moving average (SARIMA), error trend seasonality (ETS), neural network autoregression (NNAR), and Trigonometric Exponential smoothing state–space model with Box–Cox transformation, ARMA errors, Trend and Seasonal components (TBATS), and hybrid methods. These methods were applied to monthly n-FMD episodes (n = 1209) from January 2010 to December 2020. Results showed that the n-FMD episodes had a stable trend from 2010 to 2020, but they appeared to increase from 2014 to 2020. The outbreak episodes followed a seasonal pattern, with a predominant peak occurring from September to November annually. The single-technique methods yielded the best-fitting time-series models, including S A R I M A (1 , 0 , 1) (0 , 1 , 1) 12 , N N A R (3 , 1 , 2) 12 , E T S (A , N , A) , and T B A T S (1 , { 0 , 0 } , 0.8 , { < 12 , 5 > } . Moreover, SARIMA-NNAR and NNAR-TBATS were the hybrid models that performed the best on the validation datasets. The models that incorporate seasonality and a non-linear trend performed better than others. The forecasts highlighted the rising trend of n-FMD episodes in Thailand, which shares borders with several FMD endemic countries in which cross-border trading of cattle is found common. Thus, control strategies and effective measures to prevent FMD outbreaks should be strengthened not only in Thailand but also in neighboring countries. [ABSTRACT FROM AUTHOR]

Drought condition is a natural disaster that has caused economic and social damages considerably including a shortage of consuming water, and has been a hindrance to the agricultural production and industrial development. Currently, the drought condition tends to be more severe in Yasothon Province of Thailand, thus affecting plantation in the area. The purpose of this study was to monitor the drought condition of Yasothon Province, which is located in the Northeastern region of Thailand, by using Enhanced Vegetation Index (EVI) data during 2010 - 2019 obtained from Terra/Modis Satellite, and by studying the change in time series from the average EVI during 2010 - 2019, for the forecast in 2020 - 2022 by using moving averages method and exponential smoothing method in order to compare the differences between the original data of average EVI and the data of average EVI adjusted by smoothing algorithm using RMMEH method. Statistics used in examining the forecasting accuracy were MAD and MAPE. It was found from the study that MAD and MAPE of the forecast of the original average EVI and the average EVI adjusted by using RMMEH method were slightly different in that the average EVI adjusted by using RMMEH method and forecasted by moving averages method was best accurate. In addition, according to the time series change forecast of EVI, it was found that the original average EVI and the EVI which was smoothed by RMMEH method of the forecasting year during 2020 - 2022 by using moving average method and exponential smoothing method were very low in each year, indicating that the drought would occur in the future. [ABSTRACT FROM AUTHOR]

Skipjack, the most caught species of tuna globally, is a critical raw material for tuna industry in Thailand, the world's largest tuna-processing hub. However, tuna processors are finding it difficult to manage costs of these imported materials because of price fluctuations over time. Whereas most time series forecasting methods used in the literature model only three components: trend, seasonality and error, this study proposes a method to handle a fourth component as well: cycle. This method smooths monthly price data using a cubic spline that can detect cycles varying in both frequency and amplitude, and thus generates plausible forecasts by refitting the model after duplicating data from its most recent cycle. Results show that world tuna prices have a slightly upward trend in cyclical patterns with each cycle lasting approximately six years. Peak-to-peak amplitudes suggest that prices reached their peak at 2,350 US dollars per metric ton in 2017 and have started to fall, but will rebound after 2021. [ABSTRACT FROM AUTHOR]

Statistical process control methods are widely used in several fields for monitoring and detecting process problems. One of them is the control chart which is powerful and effective for monitoring many types of processes, and its capability is usually measured using the average run length (ARL). In this article, we investigate explicit formulas for both the one- and the two-sided ARL on a modified exponentially weighted moving-average (EWMA) control chart for a first-order moving-average process with exponential white noise. The accuracy of the solution obtained with the modified EWMA control chart was compared to the numerical integral equation method and extended to a comparative performance with the standard EWMA control chart. The results show that the ARL obtained by the explicit formulas and the numerical integral equation method are in close agreement. The performance comparison shows that the modified EWMA control chart is dramatically more sensitive than the standard EWMA control chart for almost all of the studied exponential smoothing parameters and magnitudes of shift size. To demonstrate its capability, the proposed approach was applied to Thailand/US monthly foreign exchange rates data, yielding in good performance. [ABSTRACT FROM AUTHOR]

The modelling of chlorophyll-a concentration helps to restrict the harmful effects in marine species caused by increased nutrient loads. The derived satellite data are often used for the monitoring of marine ecology. The common usage of satellite data is monthly average data to avoid the problem of missing values. In order to reduce the effect of missing data, this study employed the cubic spline model by using a satellite data for investigating seasonal variations of chlorophyll-a mapped in an eight-day interval consisting of missing values in Pattani Bay, Thailand from the year 2003 to 2017. This study further used the spline-fitted data for creating the baseline model of chlorophyll-a in Pattani Bay, and for examining the difference between spline fitted and monthly average data. This study revealed that the cubic spline method was able to handle the missing values in satellite data to gain the smoothness in data. When both models were compared, the spline-fitted observation yielded a smoother curve pattern than the monthly average observation. The spline fitted model was also able to display the chlorophyll-a data at any particular day of the year. It was also shown that the chlorophyll-a concentration level in the coastal area of Pattani Bay was higher in the inshore pixels, especially in rainy season. [ABSTRACT FROM AUTHOR]

In this paper, several exponential time-series methods are applied to forecast oil palm prices, crude palm oil prices, and crude palm oil production in Thailand from January to March 2018. The selected methods include Double Exponential Smoothing (DES), the Multiplicative Holt-Winters (MHW), the Additive Holt-Winters (AHW), the Improved Additive Holt-Winters (IAHW), and the Extended Additive Holt-Winters (EAHW) methods. Input data from January 2005 to December 2017 (thirteen years) were collected from the databases of the Office of Agricultural Economics and the Department of Internal Trade. Here, the novelty of our work is twofold. First, three closely related input data types are forecasted and analyzed simultaneously. Second, the well-known time-series forecasting methods (i.e. the DES, the MHW, and the AHW methods) and the efficient methods recently proposed in the literature (i.e. the IAHW and the EAHW methods) are implemented and tested. Therefore, the best forecast results determined by optimal methods are revealed. Our study demonstrates that the DES and the EAHW methods provide the smallest error (measured by Mean Absolute Percentage Error, MAPE) in forecasting oil palm and crude palm oil prices. For crude palm oil production, the IAHW and the EAHW yield better performance. The results also show the trends of the average monthly and yearly data and indicate that during January to March 2018, crude palm oil production in Thai markets should increase and prices will likely be stable. We believe that our research methodology and results can be useful for planning and setting strategy for Thai agriculturists and the government. [ABSTRACT FROM AUTHOR]

In this paper, the simple exponential smoothing (SES) and double exponential smoothing (DES) methods with designed input data are presented to forecast lime prices in Thailand during the period January 2016 to December 2016. The lime prices from January 2011 to December 2015 are the input data (i.e. seasonal data) which were gathered from the database of Simummuang market, Thailand. The major contribution of our paper is that, although, in general, the forecasting accuracy by the traditional SES and DES methods significantly decreases when those methods are used to forecast the data which show seasonality patterns, the proposed solution can properly handle such a problem. For this purpose, to forecast lime prices, 5 different input data are defined before being assigned to the SES and the DES methods: a) the monthly data of the recent year, b) the average monthly data of the past years, c) the median of the monthly data of the past years, d) the monthly data of the past years after applying the linear weighting factor, and e) the average monthly data of the past years after applying the exponential weighting factor. These designed input data are used as agents of the raw data. Our research results indicate that using the DES method with input b) and the optimal initial values to forecast lime prices during January 2016 to September 2016 significantly gives the smallest forecasting error measured by the mean absolute percentage error (MAPE). The forecast lime prices of October 2016 to December 2016 are also given. Additionally, we also demonstrate that, in our case, the SES and the DES methods with designed input data show a smaller MAPE than the methods using the multiplicative Holt-Winters and the additive Holt-Winters models which are designed for forecasting the seasonal data. [ABSTRACT FROM AUTHOR]

Abstract: The paper deals with the incidence of the Dengue Virus Infection (DVI) in the 18 districts of Sakon Nakhon Province, Thailand, from January 2005 to December 2007. Using a statistical and autoregressive analysis to smooth incidence data, we have constructed yearly and monthly district level maps of the DVI distribution. It is found that the DVI incidence is very correlated with weather conditions and higher occurrences are observed in the three most populated districts Wanon Niwat, Sawang Daen Din and Mueang Sakon Nakhon, and the virus transmission period spans from mid-summer to mid-rainy seasons (from April to August). Employing a Generalized Linear Model (GLM), we found that the DVI incidences were related with current meteorological (monthly minimum temperature, past 2-month cumulated rainfall) and socio-economical (population of 0–4years old, per capita number of public small water wells, and proportion of villages with primary schools) covariates. And using the GLM under the climate change conditions (A1B scenario of IPCC), we found that the higher risk of DVI spreads from the three most populated districts to less populated ones, and the period of virus transmission increases from 5 to 9months to include part of winter, summer and rainy seasons (from March to November) during which 6%, 61% and 33% of districts will be at low, medium and high risk of DVI occurrences, respectively. [Copyright &y& Elsevier]