GRNN神经网络python代码

GRNN（Generalized Regression Neural Network）是一种用于回归问题的神经网络模型。下面是一个简单的GRNN神经网络的Python代码示例：

import numpy as np

class GRNN:
    def __init__(self, sigma=1.0):
        self.sigma = sigma

    def gaussian_kernel(self, x, c):
        return np.exp(-np.sum((x - c) ** 2) / (2 * self.sigma ** 2))

    def train(self, X_train, y_train):
        self.X_train = X_train
        self.y_train = y_train

    def predict(self, X_test):
        y_pred = []
        for x in X_test:
            numerator = 0
            denominator = 0
            for i, c in enumerate(self.X_train):
                kernel = self.gaussian_kernel(x, c)
                numerator += kernel * self.y_train[i]
                denominator += kernel
            y_pred.append(numerator / denominator)
        return np.array(y_pred)

使用示例：

# 创建GRNN对象
grnn = GRNN(sigma=1.0)

# 训练数据
X_train = np.array([[1], [2], [3], [4], [5]])
y_train = np.array([2, 4, 6, 8, 10])
grnn.train(X_train, y_train)

# 预测数据
X_test = np.array([[6], [7], [8]])
y_pred = grnn.predict(X_test)

print(y_pred)  # 输出预测结果

这段代码实现了一个简单的GRNN神经网络模型。在训练阶段，通过调用`train`方法传入训练数据，其中`X_train`是输入特征，`y_train`是对应的目标值。在预测阶段，通过调用`predict`方法传入测试数据，即可得到预测结果。