Journal of Science and Technology, Rajabhat Maha Sarakham University

Content

2024-04-30T19:33:56+07:00

Gold Price Forecasting in Thailand using a Neural Network Model

2024-01-25T19:09:29+07:00

The purpose of this research was to create a gold price model in Thailand using an artificial neural network model under the creation of models using the CRISP-DM process, to study Factors affecting the price of gold in Thailand. Factors affecting the price of gold in Thailand were studied, to explain factors affecting the price of gold in Thailand, which has taken factors from research studies used to create a model for gold prices in Thailand, including Oil prices in the world market. Exchange rate of the baht to the United States dollar. The highest 3-month fixed deposit interest rate of commercial banks in Thailand. Inflation rate in Thailand. Thailand Consumer Confidence Index Value of imported gold bars from Thailand. Value of Thailand's successful gold exports and the Stock Exchange of Thailand stock index. Results of measuring the performance of gold price forecasting in Thailand, using two artificial neural network models. The results of the experiment found that the Model-1 model has a network architecture of 8 input layers, 4 hidden layers, and 1 node output layer by running the neural network model for 500 rounds by dividing the data into groups. A small batch size of 32, with a ratio of 90 percent modeling data and 10 percent testing, gave the best forecasting performance with a MAPE value of 8.04.

Application of FMEA to reduce wastes in the plastic injection molding industry : case study of AB Company

2024-01-13T23:36:22+07:00

This research aimed to 1) study problems occurring from the plastic injection molding process, 2) analyze and investigate the root causes of the problem and reduce waste from the plastic injection molding process, and 3) reduce production costs from the plastic injection molding process and increase productivity by reducing waste. The results showed that conducting root cause analysis using a Pareto diagram and a cause-and-effect diagram indicated that the defects were caused by black or colored spots on products in the production process. Moreover, conducting problem analysis by using the FMEA technique indicated that before the process improvement, the average RPN was 155.75 points and after the process improvement, the average RPN was 17.50 points, decreasing to 138.25 points. In addition, the main problem in the production process was insufficient cleaning of machinery and equipment. In addition, operators still did not understand the steps for cleaning and maintaining machinery and equipment which caused small powder, dust, and scraps. These could cause black or colored spots on the products in the production process. After the process improvement, the waste was reduced by 99.455 percent and the cost was reduced by 776,540 baht.

Development Free Fall Motion teaching aid with Metal Detection sensor

2024-01-13T23:34:14+07:00

This paper presents the development of a low-cost microcontroller based free-fall motion experimental devices. The free full time results was printed on LCD display. The purpose is to improve a free-fall experimental device performance by considering cost reduction, simple to use, high both accuracy and precision results. The control algorithm was implemented in an Arduino Uno R3, which is an inexpensive microcontroller unit (mcu). The height is measured from the lowest point of the metal ball to the strike plate using HC-SR04 ultrasonic sensor. To reduce the impact force on a strike plate, a method for detecting a fallen metal ball using metal detection systems. The experimental result shown that the gravity measured using the Free-Fall prototype is 10.0 m/s². Compared to the value of g measured at the nearby Udon Thani province by the National Institute of Metrology (Thailand) as 9.784156709, the value from the experiment device differs from the standard value by 2.206 %.

The Development of a Forecast Model for Predicting Product Orders Using Data Mining Techniques

2024-04-19T15:24:12+07:00

The objective of this research is to test a forecasting model for predicting product orders and determine the efficiency of the system. The researcher created using the Cross Validation 10 Folds method through data mining techniques and find satisfaction survey results of entrepreneurs who have tried using the forecasting model system to predict product orders using data mining techniques. From the 5-level score criteria using basic statistics, namely finding the average and standard deviation. It was found that the Decision Tree technique (J48) had the highest accuracy and was higher than other techniques. It is calculated as 84.84 percent, including the lowest error value, calculated as 15.16 percent, followed by the technique for searching for friends near the house (K-Nearest Neighbors), which has an accuracy value calculated as 81.60 percent, with an error value calculated in hundreds. 18.40 each and the Naïve Bayes technique has the least accuracy value, accounting for 74.11 percent, which has the highest error value, accounting for 25.89 percent.

Therefore, the Cross Validation 10 Folds method through Decision Tree technique (J48) is used to classify learning data of product orders. There were satisfaction scores regarding the development of the order forecasting model system using data mining techniques on the efficiency of the system. From a trial of 50 stores in the central region, the average was 4.12 and the standard deviation was 0.10. The interpretation score was good.

Distribution of large microplastics and mesoplastics in sand of Bangsaen and Wonnapha beaches, Chonburi province

2024-04-25T10:37:33+07:00

The objective of this study was to survey the distribution of large microplastics (size 1- <5 mm.) and mesoplastics (size 5-25 mm.) in the sand of Bangsaen and Wonnapha beaches, Chonburi province. It was a cross-sectional survey research. Sample collection was conducted in the rainy season (June 2022). The results revealed that the plastic particles occurred at every sampling site along Bangsaen and Wonnapha beaches. At Bangsaen beach, large microplastics and mesoplastics were found in the form of fragments, lines, and films. The majority of plastic particles were green. At Wonnapha beach, the shapes of large microplastics and mesoplastics were classified as EPS foam and fragment particles. The majority of plastic items were white. The findings of this research provide background data that will be useful for future monitoring and surveillance of microplastic and mesoplastic contamination.

The Association Between Health Literacy and Pregnancy Prevention Behaviors among Female Students in a University, Pathum Thani Province

2024-04-02T15:44:04+07:00

This cross-sectional descriptive research aims to study the association between health literacy and pregnancy prevention behaviors among female university students in Pathum Thani province. 407 subjects were selected by stratified sampling. The data were collected using questionnaires; including general information, cognitive health, access to health information and services, health information communication, self-management, media and information literacy, decision and pregnancy prevention behavior. The data were analyzed by descriptive statistics and inferential statistics used Pearson correlation coefficient.

The results showed that students with cognitive health 39.6% had moderate level, access to health information and services 66.8% were moderate level, health information communication 66.6% were moderate level, self-management 46.2% were moderate level, media and information literacy 40.0% were moderate level, decisions to prevent pregnancy 40.3% were in a good level and pregnancy prevention behaviors 84.8% were in a good level as well. The Pearson Correlation Coefficient analysis found that cognitive health (r=0.307) and decision (r=0.319) were low significantly and positively correlated with pregnancy prevention behavior (p<0.05). Health information communication (r=-0.161) and media and information literacy (r=-0.135) were low significantly and negatively correlated with pregnancy prevention behaviors (p<0.05). Access to health information and services and self-management, there was no correlated with pregnancy prevention behaviors. This study can be used as a basis for planning to promote the prevention of premature pregnancy by organizing activities to build confidence in preventing pregnancy for students.

Drying Kinetics and Modelling of Butterfly Pea Flower with Hot Air

2024-03-18T14:58:44+07:00

This research investigated the chemical kinetics, drying rate, specific energy consumption, and mathematical modeling of butterfly pea flower drying. The conditions of the drying tests include three levels of hot air temperature (50, 60, and 70 °C) and two levels of air velocity (1 and 2 m/s). The results showed that hot air temperature and air velocity affect the drying of butterfly pea flowers. The drying time decreased with increasing temperature and velocity. The average drying rate was 62 % at 60 °C and increased to 84% at 70 °C. The best performance for the specific energy value was drying butterfly pea flowers at 70 °C hot air temperature with 2 m/s air velocity. Hence, the value equals 16 % compared to the highest specific energy value of drying. The drying rate of butterfly pea flowers increased when the temperature and velocity increased. The results of the mathematical model prediction of the butterfly pea flower drying showed that the Wang and Singh equations were the best predictors of drying results, with R² values between 0.987970239 to 0.994963057 and χ² values between 0.006830940 to 0.009775177.

A study of bacterial growth with age of bloodstain for forensic application

2024-04-25T10:39:50+07:00

This research is a study of the relationship of bacteria growth in bloodstain with the age of the blood to use for the estimated time of the incident in forensic investigation. The experimental study was performed by comparison analysis of the bacteria counted number detected in the bloodstain from different conditions at varied duration. Blood samples were collected from a volunteer and dropped to the tested cement surface for 100 microliters and then stored in 3 conditions: indoor, outdoor and with ultraviolet source (UVC). Each test was left for 15 minutes, 30 minutes, 45 minutes, 1 hour, 2 hours, 4 hours, 8 hours, 1 day, 2 days, 3 days, 4 days, 5 days, 6 days, and 7 days after the blood was dropped. Then, numbers of bacteria were counted on the Plate Count Agar (PCA) medium using pour plate method. Data were examined and analyzed using the statistical tools of F-Test. The results revealed that the number of bacteria discovered from blood stains in indoor and natural outdoor environments did not differ substantially at the 0.05 level. When comparing the numbers of bacteria discovered under UV conditions, the results were significant at the 0.05 level.

Appendix

2024-04-30T19:36:32+07:00