Detecting heterogeneity parameters and hybrid models for precision farming

Abstract Precision farming (PF) plays a crucial role in the field of agriculture to solve challenges food shortages society. Heterogeneity, multicollinearity, and outliers are problems PF because they can cause bias lead incorrect inferences. However, traditional methods typically assume it be homogenous model, machine learning, data scientists ignore heterogeneity. In this study, aim is identify heterogeneity parameters develop hybrid models before after Data on seaweed collected using sensor smart technology attached v-Groove Hybrid Solar Drier (v-GHSD). There 29 drying parameters, each parameter has 1914 observations. We considered highest order up second interaction, increased 435 from parameters. high-dimensional data, number observations less than The authors proposed method variance inflation factor Seven predictive such as ridge, random forest, support vector machine, bagging, boosting, LASSO elastic net used select 15, 25, 35 45 significant for moisture content removal seaweed, developed robust statistical methods. For heterogeneity, model forest M Hampel with 19 best, performs better when compared other models. boosting These results vital precision farming. study will not only help us comprehend dynamics large but also gives way leverage efficient modelling.