The Results of Comparison of the Efficiency of Machine Learning Algorithms for Predicting Orchid Products in Thailand

Abstract

          This article aims to apply machine learning for predicting the orchid product in Thailand. The technique in this study was the cross-industry standard process (CRISP). There were 6 processes that were business understanding, data understanding, data preparation, model development, evaluation, and deployment. The data of planted were househole and product’s orchid collected from www.oae.go.th and www.ditp.go.th during 2016 till September 2020 that consisted of 20 provinces. To do the data preparation, feature selection for data mining i.e., data transfer, data reduction, data cleaning using Knime a tool. There were 3 algorithms for forecasting orchid product that were simple regression trees, gradient boosted tree and random forest. Algorithm performance testing by using the R² for most suitable. The results could demonstrate that the most accuracy algorithm was gradient boosted tree with accuracy 97.50% comparing with simple regression and random forest with accuracy 96.40% and 92.80%, respectively. To do the data visualization using power business intelligence (Power BI) that was very beneficial for planning and leads to higher sales.