Modern Deep Learning Design and Application Development: Versatile Tools to Solve Deep Learning Problems

Index

A

Acquisition function

.add() method

Adjusting learning rate

Adversarial learning

alter_data function

Altered task datasets

Anchor points/layers

Articulation of valued features

Asymmetric convolutions

Asymmetric factorization module

Attack performance metrics

Autoencoders

See alsoVariational autoencoder

components

data forms

denoising

SeeDenoising autoencoders (DAE)

dimensionality reduction

implementation

intuition

encoding component

feature generation

image data

convolutional autoencoder vector bottleneck design

convolutions without pooling

convolutions with pooling and padding

“enlarging” operation

“reducing” operation

shape structure and transformations

implementation

input and output

intuition and theory

pretraining

implementation

intuition

TabNet (case study)

tabular data ...