Beginning Anomaly Detection Using

padding: An integer, a tuple of two integers, a tuple of two tuples 

with two integers each. The integer tells it to add 

n rows of zeroes on 

the top and bottom of the image tensor, and 

n columns of zeroes. 

The tuple of two integers is formatted as (symmetric_height_pad, 

symmetric_width_pad), so you can tell the layer to add 

m rows of 

zeroes and 

n columns of zeroes to each side, respectively, if you pass 

in a tuple (

m, n). Finally, the tuple of two tuples is formatted as ((top_

pad, bottom_pad), (left_pad, right_pad)), so you can customize 

even more how you want the layer to add rows or columns of zeroes.

Appendix A   intro to KerAs

---

339

• 

data_format: ‘channels_first’ or ‘channels_last’. This tells the 

flattening layer how to format the flattened output to preserve the 

formatting of channels first or channels last.

 MaxPooling1D

keras.layers.MaxPooling1D()

It applies max pooling on a 1D input. To get a better idea of how max pooling works, 

check out Chapter 

3

. Max pooling in 1D is similar to max pooling in 2D, except the 

sliding window only works in one dimension, going from left to right.

This is the list of parameters:

•