训练神经网络的一般步骤

Training a Neural Network

• Randomly initialize the weights
• Implement forward propagation to get h_Θ​(x⁽ⁱ⁾) for any x⁽ⁱ⁾
• Implement the cost function
• Implement backpropagation to compute partial derivatives
• Use gradient checking to confirm that your backpropagation works. Then disable gradient checking.
• Use gradient descent or a built-in optimization function to minimize the cost function with the weights in theta.

训练一个神经网络

• 随机初始化权重
• 运用前向传播得到所有样本x⁽ⁱ⁾的h_Θ​(x⁽ⁱ⁾)
• 计算损失函数
• 运用反向传播计算偏导
• 运用梯度检查确保梯度下降算法的正确运行，然后关闭梯度检查
• 运用梯度下降算法或者别的优化算法优化权重以最小化损失函数

 

转载于:https://www.cnblogs.com/qkloveslife/p/9873681.html