Application partial least squares regression in the analysis of maize regulated deficit irrigation

In this paper, using the test-pit experiments, experimental research of the maize regulated deficit irrigation was made in black soil in cold area. The partial least-square regression was applied to set up the yield model of the maize regulated deficit irrigation, which dealed with serious multicollinearity and a small with numerous predictor variables, eliminated the bad impact of serious multicollinearity among factors, explained the dependent variables very well. The research analysis indicated that its achievement was reasonable and close to actual situation very well, providing a new idea and research method of deficit irrigation.