Simulation of daily maize evapotranspiration at different growth stages using four machine learning models in semi-humid regions of northwest China

•Construct four classical RF, SVM, ANN and ELM models.•The ELM and SVM performed slightly better than RF and ANN.•The RF and ANN models can also simulate accurately with less parameter input.•Meteorological factors and crop coefficient have significant effects on maize ETc in different growth stages.