Possible error in the concept of Pooling layer

[reubensajith]

Location Please provide: Erratum 1

1. Draft (April 30, 2021) of Machine Learning
2. Page 127
3. Second line below Figure 6.13

Describe the mistake It is mentioned as "In max pooling we instead take the average of the pixels."

Suggested change We take the average of the units in average pooling. In max pooling, we take the maximum value among the pixels within the pooling filter.

Location Please provide: Erratum 2

1. Draft (April 30, 2021) of Machine Learning
2. Page 127
3. Fifth line below Figure 6.13

Describe the mistake

It is mentioned as "to small translations of the input mkeaning if we translate the input by a small amount" **Suggested change** "to small translations of the input meaning if we translate the input by a small amount" **Location** Please provide: Erratum 3 1. Draft (April 30, 2021) of Machine Learning 2. Page 128 3. Second line under the topic "Full CNN architecture" **Describe the mistake** It is mentioned as "number of rows and columns in the hidden layers as we proceed though the network" **Suggested change** "number of rows and columns in the hidden layers as we proceed through the network"

[andreas-lindholm]

Thank you!