气象因素对香蕉批发价格影响的预测研究：以XGBoost模型为例

Predictive Study of Meteorological Factors on Banana Wholesale Price: A Case of XGBoost Model

摘要: 本研究通过集成XGBoost模型，采用主要批发市场历史价格数据和相关气象因素，开展了一项系统性的香蕉批发价格预测研究。利用参数优化及基于树的特征选择方法，实现了对模型预测精度的显著提升。结果证实了气象因素对香蕉价格有显著影响，进一步印证了我们的假设。因此，模型为市场参与者提供了准确的参考信息，有助于进行更精准的价格预测。未来的研究方向包括更深入的特征工程优化，以及探索其他可能影响香蕉价格的因素，以期进一步提高预测模型的性能和应用范围。

Abstract: This study conducted a systematic prediction of banana wholesale prices by integrating the XGBoost model with historical price data from major wholesale markets and relevant meteorological factors. The predictive accuracy of the model was significantly improved through parameter optimization and tree-based feature selection methods. The results confirmed that meteorological factors have a significant impact on banana prices, further validating our hypothesis. Therefore, the model provides accurate reference information for market participants, helping to make more precise price predictions. Future research directions include more in-depth feature engineering optimization, and exploring other factors that may affect banana prices, with the aim of further improving the predictive model's performance and applicability.