k最近邻算法:分类和回归。通过对训练集分类训练模型，验证集用于验证数据的准确性。(K nearest neighbor algorithm: classification and regression. Through the training set classification training model, the verification set is used to verify the accuracy of the data.)