Integrating Multi-Source Data for Rice Yield Prediction across China using Machine Learning and Deep Learning Approaches

•Adapting three machine/deep learning models to predict rice yield at county-level•LSTM performs better than RF and LASSO models•Combining EVI and SIF together improves yield prediction than using EVI or SIF•Develop a scalable, simple and inexpensive framework for timely predicting rice yield•The yield prediction framework use publicly available multi-source data