MACHINE LEARNING FOR SOYBEAN SEEDS LOTS CLASSIFICATION

The seed germination and vigor evaluation are essential for the sowing sector to measure performance of different lots improve efficiency storage processes. However, analysis various tests determine quality generates a large amount information, making it almost impossible humans perform quick effective control analysis. Therefore, objective this study was evaluate differences in physiological soybean seeds cultivars using machine learning techniques rank based on their quality. Three were used, germination, accelerated aging, tetrazolium treatment, seedling emergence, 1000 weight from 65 measured. evaluated two phases, one immediately after harvest other six months storage. Random forest, multi-layer perceptron, J48, classification via regression classifiers aided by feature resampler technique. forest obtained highest accuracy, random technique best results. is possible classify with great accuracy precision artificial intelligence techniques.