MODEL COMPARISONS OF CASSIFICATION FOR TOURISM DEVELOPMENT OF KHAO KHO DISTRICT, PHETCHABUN PROVINCE USING DATA MINING TECHNIQUES

Abstract

The objective of this research was to compare the classification models for tourism development in Khao Kho District, Phetchabun Province using data mining technique. The sample of this study consisted of 1,474 tourists who travel to Khao Kho District, Phetchabun Province. The sample was selected based on an accidental sampling. Data were collected between October 2018 and March 2019. A research instrument was a questionnaire compost of 7 attributes. The modeling of this research was based on seven data classification techniques, namely 1) Bayes network classifier, 2) Naive bayes, 3) Multi-layer perceptron, 4) Logistic regression, 5) Sequential minimal optimization, 6) Decision tree and 7) Random forest using RapidMiner Studio program to test the accuracy of the model from the opinions of tourists towards the development of tourist attractions in Khao Kho District, Phetchabun Province. The model performance was evaluated with k-Fold Cross Validation whereas k = 10. The results of comparing the efficiency of the model showed that the decision tree gives the highest accuracy with a value of 94.57. To compare the efficiency of the models, confusion matrix was selected from the model with the best efficiency, which was the decision tree. The experimental results could be applied to develop a predictive information modeling of tourism development in Khao Kho District, Phetchabun Province.