问题1
Which of the following do you typically see as you move to deeper layers in a ConvNet?
在典型的卷积神经网络中,随着网络的深度增加,你能看到的现象是?
- nH and nW increases, while nC also increases。nH和nW增加,nC也增加。
- nH and nW increases, while nC decreases。nH和nW增加,nC减少。
- nH and nW decrease, while nC increases。nH和nW减少,nC增加。正确
- nH and nW decreases, while nC also decreases 。nH和nW减少,nC减少。
课程链接
看图就知道
============================================================
问题2
Which of the following do you typically see in a ConvNet? (Check all that apply.)
在典型的卷积神经网络中,你能看到的是下面哪些内容?
- Multiple CONV layers followed by a POOL layer。多个卷积层后面跟着的是一个池化层。正确。
- Multiple POOL layers followed by a CONV layer。多个池化层后面跟着的是一个卷积层。
- FC layers in the last few layers。全连接层(FC)位于最后的几层。正确。
- FC layers in the first few layers。全连接层(FC)位于开始的几层。
课程链接
看图就知道
============================================================
问题3
In order to be able to build very deep networks, we usually only use pooling layers to downsize the height/width of the activation volumes while convolutions are used with “valid” padding. Otherwise, we would downsize the input of the model too quickly.
为了能够构建非常深的网络,我们通常只使用池层来缩小激活卷的高度/宽度,卷积使用“valid”填充。否则,我们将过快地缩小模型的输入。
- True。
- False。正确
课程链接
参见课程中VGG网络,使用的是same填充。
============================================================
问题4
Training a deeper network (for example, adding additional layers to the network) allows the network to fit more complex functions and thus almost always results in lower training error. For this question, assume we’re referring to “plain” networks.
假设我们使用一个普通网络(结构)来训练一个更深的网络(例如,向网络中添加额外的层)可以使网络适应更复杂的功能,因此几乎总是可以降低训练误差。
- True。
- False。正确。
课程链接
课程中有普通网络和残差网络对比。
在有残差的ResNet中,即使网络再深,训练误差都会随着网络层数的加深逐渐减小。这是残差网络的优点。
============================================================
问题5
The following equation captures the computation in a ResNet block. What goes into the two blanks above?
下面计算残差块的公式中,横线上应该分别填什么?
- 0 and z [ l + 1 ] z^{[l+1]} z[l+1], respectively。分别是0 和 z [ l + 1 ] z^{[l+1]} z[l+1]
- a [ l ] a^{[l]} a[l] and 0, respectively。分别是 a [ l ] a^{[l]} a[l] 和 0。正确
- 0 and a [ l ] a^{[l]} a[l], respectively。分别是0 和 a [ l ] a^{[l]} a[l]
- z [ l ] z^{[l]} z