Forecasting the Main Energy Crop Prices in the Agricultural Sector of Thailand Using a Machine Learning Model

This study uses a machine learning algorithm to forecast the price of Thailand’s primary energy crops. The information is annually compiled from 2000 to 2022 and covers the prices of the five primary energy crops used in the agricultural industry. Empirically, two regimes—the booming market and the recessionary market—are offered by the unsupervised learning k-means method, which is used to cluster the cycle regimes of agricultural prices. Seven years indicate the beginning of a rising market, and 15 years mark the beginning of a recessionary one. Furthermore, the cycle regimes of energy crop prices in the upcoming 5 years are investigated utilizing supervised learning techniques such as linear discriminant analysis (LDA), k-nearest neighbors (kNN), and support vector machines (SVMs). The findings showed that LDA was selected based on the greatest coefficient validation, representing crop price regimes in Thailand’s agricultural industry that will continue to experience recessionary periods over the following 5 years. Therefore, the knowledge presented in this chapter might help farmers control their production, especially in the domain of energy crops. To promote the production and use of energy crops, it is essential to maintain stable and competitive prices. Policymakers can support the energy crop industry by providing incentives to farmers, investing in research and development, and promoting the adoption of biofuels in the transportation sector. By doing so, Thailand can enhance its energy security, promote sustainable agriculture, and support economic growth in a recessionary market.