Intelligence System via Machine Learning Algorithms in Detecting the Moisture Content Removal Parameters of Seaweed Big Data

Abstract

The parameters that determine the removal of moisture content have become necessary in seaweed research as they can reduce cost and improve the quality and quantity of the seaweed. During the seaweed’s drying process, many drying parameters are involved, so it is hard to find a model that can determine the drying parameters. This study compares seaweed big data performance using machine learning algorithms. To achieve the objectives, four machine learning algorithms, such as bagging, boosting, support vector machine, and random forest, were used to determine the significant parameters from the data obtained from v-GHSD (v-Groove Hybrid Solar Drier). The mean absolute percentage error (MAPE) and coefficient of determination (R2) were used to assess the model. The importance of variable selection cannot be overstated in big data due to the large number of variables and parameters that exceed the number of observations. It will reduce the complexity of the model, avoid the curse of dimensionality, reduce cost, remove irrelevant variables, and increase precision. A total of 435 drying parameters determined the moisture content removal, and each algorithm was used to select 15, 25, 35 and 45 significant parameters. The MAPE and R-Square for the 45 highest variable importance for random forest are 2.13 and 0.9732, respectively. It performed best, with the lowest error and the highest R-square. These results show that random forest is the best algorithm to decide the vital drying parameters for removing moisture content.