The food problem is a major issue of common concern around the world, and the forecasting of wheat output can help to promote the solution of this problem. This paper analyses the relationship between wheat yield and weather, precipitation and irrigated area; proposes a Bayesian optimized machine model of the XGBoost regression algorithm; and applies it to forecast wheat yield in Uttar Pradesh, India. Second, the stability and robustness of the XGBoost model using the top five parameter combinations of the optimization score were evaluated by cross-validation, and the best model was used to predict wheat yield in Uttar Pradesh, India. Finally, the model is compared with other prediction models, and the tenfold cross-validation experimental results show the effectiveness of the model. Ultimately, by learning from the available data and using the 2022 wheat irrigation data to forecast wheat yield per acre, we obtain a wheat production estimate of 17 930 817 t in Uttar Pradesh in 2022, which is a decreasing trend from previous years' wheat output. The impact of extreme heat on total production is somewhat contained by the year-on-year increase in the irrigated area in Uttar Pradesh.