PREDICTION OF BOESENBERGIA ROTUNDA (L.) DRYING USING HOT AIR AND ULTRASONIC VIBRATION VIA EXTREME GRADIENT BOOSTING
การทำนายการอบแห้งของกระชายขาวด้วยลมร้อนและการสั่นสะเทือนความถี่สูงโดยใช้เทคนิคเอ็กซ์ตรีมเกรเดียนต์บูสติง
DOI : https://doi.org/10.35633/inmateh-76-43
Authors
Abstract
This study employed gradient boosting based machine learning (ML) models to investigate white fingerroot drying by hot air with ultrasonic vibration. The dataset consisted of 73 data points, with three input features namely temperature, time and vibration, and one target output: average moisture, posing a significant challenge for developing ML models. In this study, the dataset was split into 75% training data and 25% test data to evaluate the performance of the ML model. The model showed high prediction accuracy on the test data, achieving an R² value of 0.99 and a RMSE of 8.71. However, due to the small dataset, the training data yielded a slightly lower accuracy, with an R² value of 0.96 and an RMSE of 25.66. An analysis was also conducted to explain how individual variables influenced the model's predictions. Using SHAP, the relationship between vibration and drying time was examined, revealing that vibration had a more positive effect on average moisture content during longer drying time.
Abstract in Thai