Data augmentation
Data augmentation is the process of adding to or transforming data to increase its variety or usefulness for learning, computation, or analysis.
Deep Learning with R, Third Edition
Data augmentation is a technique that generates additional training examples by applying label-preserving transformations to existing data.
- How it is used
- Mitigates overfitting when training image classifiers on small datasets.
- How deep it goes
- Explains random image transformations, pipeline placement, performance tradeoffs, and training with augmented batches.
- Examples
- Uses cats-versus-dogs images to demonstrate random flips, rotations, zooms, and test-set evaluation.
- What you can do after
- Apply it using R and Keras augmentation layers in a training data pipeline.
Read how Deep Learning with R, Third Edition explains data augmentation →
Deep Learning with PyTorch, Second Edition
A training technique that creates modified versions of examples to increase effective dataset diversity while preserving task-relevant meaning.
- How it is used
- Uses synthetic sample alterations to combat overfitting when training data is limited.
- How deep it goes
- Works through configurable CT transformations, affine resampling, caching order, and validation comparisons.
- Examples
- Compares mirrored faces with randomized corner pixels and applies flips, shifts, scaling, rotation, and noise to CT nodule samples.
- What you can do after
- Apply it via PyTorch by configuring CT transformations and comparing validation runs in TensorBoard.
Read how Deep Learning with PyTorch, Second Edition explains data augmentation →
Deep Learning with Python, Third Edition
A regularization approach that creates additional training examples by applying label-preserving random transformations to existing samples.
- How it is used
- Uses data augmentation to mitigate overfitting when training image classifiers on small datasets.
- How deep it goes
- Demonstrates transformation choices, pipeline placement, visualization, and training; notes that augmented samples remain correlated.
- Examples
- Uses random flips, rotations, and zooms on training images, then trains a ConvNet with dropout.
- What you can do after
- Apply it via Keras augmentation layers and a TensorFlow dataset pipeline, then inspect the transformed images.
Read how Deep Learning with Python, Third Edition explains data augmentation →
Delivering Machine Learning Projects: From design to deployment
Data augmentation is the creation of additional training examples by applying transformations or alterations to existing data.
- How it is used
- Uses transformations and alterations to expand narrow training sets and support more robust models.
- How deep it goes
- Explains the rationale and lists image and text transformations, while leaving implementation details of each manipulation unspecified.
- Examples
- Uses rotated, mirrored, noisy, scaled, repositioned, and otherwise altered cat images, plus spelling, synonym, and translation changes for text.
- What you can do after
- Build additional training data that can help models become more robust when original examples are narrow.
Human-in-the-Loop Machine Learning: Active learning and annotation for human-centered AI
A technique that creates modified or additional data examples to increase the amount or diversity of data available for learning.
- How it is used
- Expands training data diversity when important examples are scarce and supports models that aggregate object annotations.
- How deep it goes
- Explains image and language transformations with examples, while deferring algorithm-focused computer vision instruction.
- Examples
- Illustrates back translation with positive sentiment and masked-word prediction with variations of a sentence about a trip to a shop.
- What you can do after
- Evaluate augmentation options for image or text data and understand that generated examples may not always retain their labels.
Machine Learning for Tabular Data: XGBoost, Deep Learning, and AI
A technique that expands or diversifies training data by applying label-preserving transformations or generating additional examples.
- How it is used
- Mitigates deep learning’s difficulty with small training datasets.
- How deep it goes
- Defines augmentation’s purpose but stops short of describing transformations, implementation steps, or evaluation procedures.
- Examples
- Uses the Tokyo and New York City Airbnb datasets to illustrate contrasting data volumes and deep-learning challenges.
- What you can do after
- Evaluate when augmentation may help a deep-learning model facing limited training data.
Deep Learning with JAX
The practice of creating modified training examples to increase diversity and improve model robustness.
- How it is used
- Uses randomized image transformations to increase training diversity, improve robustness, and reduce overfitting.
- How deep it goes
- Explains augmentation through two image transformations and connects it to implicit regularization and random-number handling.
- Examples
- Demonstrates image classification with random noise and random rotations applied to training images.
- What you can do after
- Apply data augmentation with JAX's functional random-number model to diversify image-classification training data.
Read how Deep Learning with JAX explains data augmentation →
Build AI Drug Discovery Pipelines
Data augmentation is the creation or transformation of training examples to increase the effective diversity of a dataset.
- How it is used
- Uses transformed training examples to improve robustness and address repetitive-instance problems in imbalanced classification.
- How deep it goes
- Defines augmentation, explains its regularization rationale, and gives image and molecular transformation examples without prescribing a full implementation.
- Examples
- Demonstrates image rotations, flips, and zooms alongside molecular SMILES edits involving functional groups, atoms, and bonds.
- What you can do after
- Apply transformations to create varied training instances and reduce memorization during model training.
Read how Build AI Drug Discovery Pipelines explains data augmentation →
Inside Deep Learning: Math, Algorithms, Models
The practice of creating additional training examples by applying label-preserving transformations to existing data.
- How it is used
- Augments CNN training data with label-preserving image transformations to increase input diversity when real data is limited.
- How deep it goes
- Defines randomized transforms, rotation ranges, application probabilities, and visual checks, while leaving transform choice data-dependent.
- Examples
- Demonstrates affine changes, crops, flips, perspective, color jitter, and MNIST flip failures.
- What you can do after
- Apply torchvision.transforms to augment training images and inspect whether transformed labels remain recognizable.
Read how Inside Deep Learning: Math, Algorithms, Models explains data augmentation →
Generative AI in Action
Data augmentation is the creation or transformation of training examples to increase dataset variety while preserving task-relevant information.
- How it is used
- Uses synthetic visual data to cover edge cases and insufficient diversity in training datasets.
- How deep it goes
- Explains GAN-based synthetic-data augmentation and names mode collapse, instability, output control, and bias risks without implementation steps.
- Examples
- Uses education training simulations, rare edge cases, and datasets lacking sufficient diversity as augmentation scenarios.
- What you can do after
- Apply it via GANs to create synthetic training data for edge cases or scarce, insufficiently diverse datasets.
Read how Generative AI in Action explains data augmentation →
Sutskever's List
Data augmentation is the creation of varied training examples by applying label-preserving transformations to existing data.
- How it is used
- Expands the effective training set and supports generalization by exposing models to realistic variation.
- How deep it goes
- Distinguishes geometric transforms from PCA-based color jitter and their purposes, without specifying implementation parameters.
- Examples
- Uses random crops, horizontal reflections, and PCA-based color jitter on ImageNet training images.
- What you can do after
- Understand: Explain how geometric and color variations can support generalization in image classification.
Timeless Algorithms, The Foundational Ideas
Data augmentation creates additional training observations by transforming existing observations while preserving their correct outcomes.
- How it is used
- Constrains how a network uses its capacity to support generalization.
- How deep it goes
- Defines label-preserving transformations and explains their role in learning predictive patterns across changing conditions.
- Examples
- Uses pedestrian images transformed by cropping, translation, reflection, or brightness changes.
- What you can do after
- Understand how label-preserving variation can discourage memorization of particular training examples.
Read how Timeless Algorithms, The Foundational Ideas explains data augmentation →
Also covered in
- Snowflake Data Engineering
- Grokking Simplicity
- Learn Generative AI with PyTorch
- Build a Text-to-Image Generator (from Scratch)
- Machines that ThinkIn early access (MEAP)
What to know first
- overfittingOverfitting is a modeling condition in which a model adapts too closely to particular data, including its noise or idiosyncrasies, and consequently generalizes poorly to new data.
- classifierA classifier is a predictive model that assigns an input to one or more discrete classes or categories.
- accuracyAccuracy is the degree to which a result agrees with a specified truth, reference, or criterion, commonly measured as the proportion of evaluated predictions that are correct.
- computer visionComputer vision is the field of computing concerned with extracting and interpreting information from images, video, and other visual data.
Related concepts
- AlexNetused by · AlexNet is a convolutional neural network architecture designed for image classification and recognized for its landmark performance in large-scale image recognition.
- overfittingaddresses · Overfitting is a modeling condition in which a model adapts too closely to particular data, including its noise or idiosyncrasies, and consequently generalizes poorly to new data.
- training datasetproduces · A training dataset is a collection of examples supplied to a learning procedure to fit or adapt a model’s parameters or behavior.
- demonstrationuses · A demonstration is an example, often pairing an input with an output, that illustrates a task or desired behavior for training or guiding a model.
Questions about data augmentation
What is data augmentation?
Data augmentation is the process of adding to or transforming data to increase its variety or usefulness for learning, computation, or analysis.
Which Manning books explain data augmentation?
Deep Learning with R, Third Edition, Deep Learning with PyTorch, Second Edition, Deep Learning with Python, Third Edition, Delivering Machine Learning Projects: From design to deployment, Human-in-the-Loop Machine Learning: Active learning and annotation for human-centered AI, Machine Learning for Tabular Data: XGBoost, Deep Learning, and AI, Deep Learning with JAX, Build AI Drug Discovery Pipelines and 4 more.
What should I know before learning data augmentation?
Manning books that teach data augmentation usually introduce overfitting, classifier, accuracy and computer vision first.
What kinds of data augmentation are there?
Kinds of data augmentation include back translation.
What is data augmentation also called?
Data augmentation is also called dataset augmentation.
These guides are generated from the text of Manning books by Manning's knowledge graph and link to the sections they draw on. The full books are available on liveBook and at manning.com.











