AGRARIAN/vineyards-yield-dataset

该数据集用于训练机器学习模型，以预测葡萄收获时的产量（TCH）。数据集包含926个样本，每个样本有589个特征，这些特征包括卫星数据（如NDVI、EVI、VARI、NDRE、NDWI等）、天气时间序列（如降水、温度、度日等）、土壤属性（如粘土、沙子、氮含量等）以及农艺属性（如品种、年龄、切割周期等）。数据集通过Leave-One-Season-Out Cross-Validation方法进行模型训练和评估，最终生成用于预测的模型文件。

This dataset is used to train machine learning models for predicting vine yield (TCH) at harvest time. The dataset contains 926 samples, each with 589 features, including satellite data (e.g., NDVI, EVI, VARI, NDRE, NDWI), weather time series (e.g., precipitation, temperature, degree days), soil properties (e.g., clay, sand, nitrogen), and agronomic attributes (e.g., variety, age, cut cycle). The dataset is used for model training and evaluation via Leave-One-Season-Out Cross-Validation, ultimately producing model files for prediction.