KMITL

Permanent URI for this communityhttps://dspace.kmitl.ac.th/handle/123456789/1

Browse

Search Results

Now showing 1 - 10 of 21

Some of the metrics are blocked by your
consent settings
Comparative Evaluation of Event-Based Forecasting Models for Thai Airport Passenger Traffic
(2026-02-01)
Chaikajonwat, Thanrada
;
Araveeporn, Autcha
Accurate passenger traffic forecasting is vital for strategic planning in Thailand’s aviation industry. This study forecasts the monthly total number of passengers at Suvarnabhumi (BKK), Don Mueang (DMK), Chiang Mai (CNX), and Phuket (HKT) airports using data from 2017 to 2024. The dataset was partitioned into training (January 2017–December 2023) and testing (January–December 2024) sets. Six methods were compared: Single Exponential Smoothing, Holt’s, Holt’s with Events Adjustment, Holt–Winters Multiplicative, TBATS model, and Box–Jenkins. Performance was evaluated using Mean Absolute Percentage Error (MAPE) and Mean Absolute Error (MAE). The results indicate that the optimal forecasting method varies by airport characteristics. Holt’s Method with Events Adjustment, which incorporates major disruptions such as the COVID-19 pandemic, produced the most accurate forecasts for BKK and DMK by effectively capturing external shocks. In contrast, the Holt–Winters Multiplicative method performed best for CNX and HKT, reflecting strong seasonal patterns typically driven by tourism activities in these destinations.
Some of the metrics are blocked by your
consent settings
An Enhanced Discriminant Analysis Approach for Multi-Classification with Integrated Machine Learning-Based Missing Data Imputation
(2025-11-01)
Araveeporn, Autcha
;
Kangtunyakarn, Atid
This study addresses the challenge of accurate classification under missing data conditions by integrating multiple imputation strategies with discriminant analysis frameworks. The proposed approach evaluates six imputation methods (Mean, Regression, KNN, Random Forest, Bagged Trees, MissRanger) across several discriminant techniques. Simulation scenarios varied in sample size, predictor dimensionality, and correlation structure, while the real-world application employed the Cirrhosis Prediction Dataset. The results consistently demonstrate that ensemble-based imputations, particularly regression, KNN, and MissRanger, outperform simpler approaches by preserving multivariate structure, especially in high-dimensional and highly correlated settings. MissRanger yielded the highest classification accuracy across most discriminant analysis methods in both simulated and real data, with performance gains most pronounced when combined with flexible or regularized classifiers. Regression imputation showed notable improvements under low correlation, aligning with the theoretical benefits of shrinkage-based covariance estimation. Across all methods, larger sample sizes and high correlation enhanced classification accuracy by improving parameter stability and imputation precision.
Some of the metrics are blocked by your
consent settings
Improved Probability-Weighted Moments and Two-Stage Order Statistics Methods of Generalized Extreme Value Distribution
(2025-07-01)
Araveeporn, Autcha
This study evaluates six parameter estimation methods for the generalized extreme value (GEV) distribution: maximum likelihood estimation (MLE), two probability-weighted moments (PWM-UE and PWM-PP), and three robust two-stage order statistics estimators (TSOS-ME, TSOS-LMS, and TSOS-LTS). Their performance was assessed using simulation experiments under varying tail behaviors, represented by three types of GEV distributions: Weibull (short-tailed), Gumbel (light-tailed), and Fréchet (heavy-tailed) distributions, based on the mean squared error (MSE) and mean absolute percentage error (MAPE). The results showed that TSOS-LTS consistently achieved the lowest MSE and MAPE, indicating high robustness and forecasting accuracy, particularly for short-tailed distributions. Notably, PWM-PP performed well for the light-tailed distribution, providing accurate and efficient estimates in this specific setting. For heavy-tailed distributions, TSOS-LTS exhibited superior estimation accuracy, while PWM-PP showed a better predictive performance in terms of MAPE. The methods were further applied to real-world monthly maximum PM2.5 data from three air quality stations in Bangkok. TSOS-LTS again demonstrated superior performance, especially at Thon Buri station. This research highlights the importance of tailoring estimation techniques to the distribution’s tail behavior and supports the use of robust approaches for modeling environmental extremes.
Some of the metrics are blocked by your
consent settings
Parameter Estimation For Generalized Extreme Value Distribution In Rainfall Forecasting: A Case Study Of Bangkok
(2025-01-01)
Araveeporn, Autcha
;
Sukpan, Paradorn
This study aims to compare efficiency methods for the estimated parameter of Generalized Extreme Value Distribution (GEVD), which consists of location, scale, and shape parameters. The parameter estimation employs the maximum likelihood (ML), generalized maximum likelihood (GML), Bayesian, and L-moments methods. The data is generated through simulation data in Gumbel, Fréchet, and Weibull distributions verified by shape parameters. The performance of these methods is evaluated using the minimum mean squared error (MSE) and mean absolute percentage error (MAPE). The results indicate that the Bayesian, ML, and GML methods consistently achieve the lowest MSE values, such as 0.0120 for location, 0.0066 for scale, and 0.005 for shape parameters when the sample size is 100 of the Gumbel distribution. In the real data application using 29 years of Bangkok rainfall data (1994-2023), the GEVD model estimated return levels for 2 to 9 years, with MAPE values ranging from 32.57% to 56.92% across different stations. The findings suggest that ML and GML methods outperform others in simulated and real-world applications. The proposed approach provides accurate and reliable forecasts of extreme rainfall, which are crucial for Bangkok’s urban planning and flood risk management.
Some of the metrics are blocked by your
consent settings
The Use of Spline Techniques in the Nonparametric Regression Analysis for the Sequence Data with a Random Walk Process
(2025-01-01)
Araveeporn, Autcha
;
Chaikajonwat, Thanrada
This study evaluates and compares various spline techniques in the nonparametric regression analysis, specifically focusing on the smoothing spline regression, the natural spline regression, the B-spline regression, and the penalized spline regression. The dependent variable in this analysis is time series data generated by a random walk process, while the independent variable is represented as sequential data. The simulation data, derived from a random walk process with diverse variances and sample sizes, ensures an absence of fixed patterns in the variable's changes. In addition, real-world data from the monthly trading volume of the SET (Stock Exchange of Thailand) index is used for practical application. The criterion for model efficiency estimation is based on minimizing the average mean square error for the simulation and SET index data. At the same time, predictive performance for future values is assessed through the minimum of average mean absolute percentage error. Among the models tested, the natural spline regression achieved the minimum average mean square error in all simulations due to SET index data estimation, excelling in model fit. However, the B-spline regression proved highly effective for forecasting future values.
Some of the metrics are blocked by your
consent settings
Adaptive Multiclassification With Lung Cancer Types Using High-Dimensional Discriminant Analysis and Machine Learning Methods
(2025-01-01)
Araveeporn, Autcha
This research investigates adaptive multiclassification methods for classifying lung cancer types using both high-dimensional discriminant analysis (HDDA) and widely used machine learning (ML) approaches under challenging data conditions, including high dimensionality, multicollinearity, outliers, and imbalanced classes. The dataset consists of 1000 gene expressions as explanatory variables and four lung cancer types as response variables, categorizing the problem as high-dimensional and imbalanced. HDDA introduces a statistically principled parametrization of the covariance matrix tailored for high-dimensional data. At the same time, ML methods such as Naïve Bayes, K-Nearest Neighbors, Support Vector Machine, Artificial Neural Network, and Random Forest offer flexible, data-driven alternatives. While previous studies have separately investigated either discriminant analysis or ML, there is a lack of comparative studies that evaluate their performance simultaneously under such complex conditions. This study addresses this gap by systematically analyzing both approaches with balanced and imbalanced gene expression data. The central hypothesis of this study is that HDDA, despite being a classical statistical technique, can achieve performance comparable to or complementary with ML methods when applied to gene expression data. Experiments on both original and balanced datasets, across varying subsets of explanatory variables, show that data balancing consistently improves accuracy, precision, and recall. Among ML methods, Random Forest achieves the highest predictive performance on balanced data, while HDDA provides competitive and interpretable results across scenarios.
Some of the metrics are blocked by your
consent settings
Forecasting Models for Total Crude Palm Oil Productions in Thailand
(2024-12-02)
Banditvilai, Somsri
;
Araveeporn, Autcha
This research aims to find a suitable forecasting model for Thailand's total crude palm oil production. The monthly total crude palm oil production in Thailand was gathered from the Office of Agricultural Economics, Ministry of Agriculture, and cooperatives from January 2010 to December 2022. The data were divided into two sets. The first set, from January 2010 to December 2021, was used for constructing and selecting the forecasting models. The second one, from January 2022 to December 2022, was used to compute the accuracy of the forecasting model. Since the total crude palm oil production has trend and seasonal variation, the research used the Holt-Winters method with different initial settings for trend and seasonal influence, the Bagging Holt-Winters method, and the Box-Jenkins method to construct the forecasting models. The minimum mean square error (MSE) and residuals have normal distributions used to select the appropriate forecasting model, and the mean absolute percentage error (MAPE) was used to compute the efficiency of the forecasting model.According to the three forecasting methods results, the Box-Jenkins method was suitable for forecasting Thailand's total crude palm oil production. The ARIMA(2,1,2)(0,1,1)12 model was the best model for predicting Thailand's total crude palm oil production and yielded the MAPE =13.49%
Some of the metrics are blocked by your
consent settings
Comparison of Machine Learning Methods for Binary Classification of Multicollinearity Data
(2024-12-02)
Araveeporn, Autcha
;
Wanitjirattikal, Puntipa
This study examines the effectiveness of binary classification performance in multicollinearity. Four machine learning methods, namely backpropagation neural network, Naïve Bayes, support vector machine, and random forest, are compared in terms of their efficiency in handling multicollinear data. The evaluation of binary classification performance efficiency considers multicollinearity in independent variables, considering both a constant correlation model and the Toeplitz correlation. Correlation coefficients of 0.1 and 0.9 are explored in the analysis. The independent variables in this study are simulated from a multivariate normal distribution with 10, 20, 30, and 40 variables, respectively. The dependent variable is constructed using the logit function with sample sizes of 100 and 200. The simulation and data analysis are performed using the R Studio program and repeated 1,000 times for each scenario. The findings of this research reveal that the backpropagation neural network and Naïve Bayes methods exhibit superior performance in determining the mean accuracy percentage under constant correlation. On the other hand, the backpropagation neural network and support vector machine are the most effective methods in determining the mean accuracy percentage when dealing with multicollinearity in the form of Toeplitz correlation.
Some of the metrics are blocked by your
consent settings
The Estimating Parameter and Number of Knots for Nonparametric Regression Methods in Modelling Time Series Data
(2024-12-01)
Araveeporn, Autcha
This research aims to explore and compare several nonparametric regression techniques, including smoothing splines, natural cubic splines, B-splines, and penalized spline methods. The focus is on estimating parameters and determining the optimal number of knots to forecast cyclic and nonlinear patterns, applying these methods to simulated and real-world datasets, such as Thailand’s coal import data. Cross-validation techniques are used to control and specify the number of knots, ensuring the curve fits the data points accurately. The study applies nonparametric regression to forecast time series data with cyclic patterns and nonlinear forms in the dependent variable, treating the independent variable as sequential data. Simulated data featuring cyclical patterns resembling economic cycles and nonlinear data with complex equations to capture variable interactions are used for experimentation. These simulations include variations in standard deviations and sample sizes. The evaluation criterion for the simulated data is the minimum average mean square error (MSE), which indicates the most efficient parameter estimation. For the real data, monthly coal import data from Thailand is used to estimate the parameters of the nonparametric regression model, with the MSE as the evaluation metric. The performance of these techniques is also assessed in forecasting future values, where the mean absolute percentage error (MAPE) is calculated. Among the methods, the natural cubic spline consistently yields the lowest average mean square error across all standard deviations and sample sizes in the simulated data. While the natural cubic spline excels in parameter estimation, B-splines show strong performance in forecasting future values.
Some of the metrics are blocked by your
consent settings
Empirical Comparison of Forecasting Methods for Air Travel and Export Data in Thailand
(2024-12-01)
Banditvilai, Somsri
;
Araveeporn, Autcha
Time series forecasting plays a critical role in business planning by offering insights for a competitive advantage. This study compared three forecasting methods: the Holt–Winters, Bagging Holt–Winters, and Box–Jenkins methods. Ten datasets exhibiting linear and non-linear trends and clear and ambiguous seasonal patterns were selected for analysis. The Holt–Winters method was tested using seven initial configurations, while the Bagging Holt–Winters and Box–Jenkins methods were also evaluated. The model performance was assessed using the Root-Mean-Square Error (RMSE) to identify the most effective model, with the Mean Absolute Percentage Error (MAPE) used to gauge the accuracy. Findings indicate that the Bagging Holt–Winters method consistently outperformed the other methods across all the datasets. It effectively handles linear and non-linear trends and clear and ambiguous seasonal patterns. Moreover, the seventh initial configurationdelivered the most accurate forecasts for the Holt–Winters method and is recommended as the optimal starting point.

KMITL

Browse

Filters

Settings

Sort By

Results per page

Search Results