Machine Learning Mapping of Soil Apparent Electrical Conductivity on a Research Farm in Mississippi

Open-source and free tools are readily available to the public process data assist producers in making management decisions related agricultural landscapes. On-the-go soil sensors being used as a proxy develop digital maps because of they can collect their ability cover large area quickly. Machine learning, subcomponent artificial intelligence, makes predictions from data. Intermixing open-source tools, on-the-go sensor technologies, machine learning may improve Mississippi mapping crop production. This study aimed evaluate for apparent electrical conductivity (ECa) collected with an system at two sites (i.e., MF2, MF9) on research farm Mississippi. (support vector machine) incorporated Smart-Map, application, were derive maps. Autocorrelation shallow (ECas) deep (ECad) readings was statistically significant both locations (Moran’s I, p 0.001); however, spatial correlation greater MF2. According leave-one-out cross-validation results, best models developed ECas versus ECad. Spatial patterns observed ECad fields. The more distinct than measurements. results indicated that valuable deriving Location depth played role learner’s