Abstract: This paper explores how high-throughput phenotyping can be integrated with machine learning models to efficiently forecast the Sterility Mosaic Disease using a small amount of training data.