Beginning Anomaly Detection Using

activity_regularizer: A regularizer function applied to the output of 

the layer.

• 

kernel_constraint: A constraint function applied to the weights. For 

more information, check out the 

Constraints section.

• 

bias_constraint: A constraint function applied to the bias.

Appendix A   intro to KerAs

---

336

 Conv2D

keras.layers.Conv1D()

Check out Chapter 

3

 for a detailed explanation on how the 2D convolutional layer 

works.

This layer is a two-dimensional convolutional layer. It basically passes a 2D filter over 

the input and multiplies the values element-wise to create the output feature map.

These are the parameters that the function takes:

• 

filters: An integer value that determines the dimensionality of the 

output space. In other words, this is also the number of filters in the 

convolution.

• 

kernel_size: An integer (or tuple/list of two integers) that specifies 

the height and width of the filter/kernel that is used in the 2D 

convolution.

• 

strides: An integer (or tuple/list of two integers, one for height and 

one for width, respectively) that tells the layer how many data entries 

to shift by after one element-wise multiplication of the filter and the 

input data. Note: A stride value != 1 isn’t compatible if the dilation_

rate != 1.

• 

padding: ‘valid’ or ‘same’. ‘valid’ doesn’t zero pad the output. ‘same’ 

zero pads the output so that it’s the same length as the input.

• 

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.

• 

dilation_rate: An integer (or tuple/list of a two integers) serves as the 

dilation rate for this dilated convolutional layer. For an explanation of 

how this works, refer to Chapter 

7

.

•