Neural Networks and Deep Learning 课程笔记(第四周)深层神经网络(Deep Neural Networks)

时间:2020-12-23 07:08:10

1. 深层神经网络(Deep L-layer neural network )

Neural Networks and Deep Learning 课程笔记(第四周)深层神经网络(Deep Neural Networks)

Neural Networks and Deep Learning 课程笔记(第四周)深层神经网络(Deep Neural Networks)

2. 前向传播和反向传播(Forward and backward propagation)

Neural Networks and Deep Learning 课程笔记(第四周)深层神经网络(Deep Neural Networks)

Neural Networks and Deep Learning 课程笔记(第四周)深层神经网络(Deep Neural Networks)

Neural Networks and Deep Learning 课程笔记(第四周)深层神经网络(Deep Neural Networks)

Neural Networks and Deep Learning 课程笔记(第四周)深层神经网络(Deep Neural Networks)

Neural Networks and Deep Learning 课程笔记(第四周)深层神经网络(Deep Neural Networks)

3. 总结

Neural Networks and Deep Learning 课程笔记(第四周)深层神经网络(Deep Neural Networks)

Neural Networks and Deep Learning 课程笔记(第四周)深层神经网络(Deep Neural Networks)

4. 深层网络中的前向传播(Forward propagation in a Deep Network)

Neural Networks and Deep Learning 课程笔记(第四周)深层神经网络(Deep Neural Networks)

Neural Networks and Deep Learning 课程笔记(第四周)深层神经网络(Deep Neural Networks)

向量化实现过程可以写成:

Neural Networks and Deep Learning 课程笔记(第四周)深层神经网络(Deep Neural Networks)

注:这里只能用一个显示for循环,l 从 1 到 L,然后一层接着一层去计算。

如何减少bug

4.1 核对矩阵的维数(Getting your matrix dimensions right)

Neural Networks and Deep Learning 课程笔记(第四周)深层神经网络(Deep Neural Networks)

Neural Networks and Deep Learning 课程笔记(第四周)深层神经网络(Deep Neural Networks)

4.2 向量化实现

Neural Networks and Deep Learning 课程笔记(第四周)深层神经网络(Deep Neural Networks)

Neural Networks and Deep Learning 课程笔记(第四周)深层神经网络(Deep Neural Networks)

4.3 为什么使用深层表示?(Why deep representations? )

4.4 搭建神经网络块(Building blocks of deep neural networks)

Neural Networks and Deep Learning 课程笔记(第四周)深层神经网络(Deep Neural Networks)

4.5 参数 VS  超参数(Parameters vs Hyperparameters )

Neural Networks and Deep Learning 课程笔记(第四周)深层神经网络(Deep Neural Networks)

4.6 总结

Neural Networks and Deep Learning 课程笔记(第四周)深层神经网络(Deep Neural Networks)