Representation learning
Representation learning is the process by which a machine-learning model learns useful representations or features of data from examples.
Machine Learning for Tabular Data: XGBoost, Deep Learning, and AI
Representation learning is the automatic discovery of useful data representations from input data.
- How it is used
- Frames deep learning as an approach that discovers task-relevant data representations internally.
- How deep it goes
- Explains nonlinear interactions, compact formats, and hierarchical representations without detailing specific network architectures.
- Examples
- Uses hierarchical transformations from basic features to complex abstract patterns and compact task-specific formats.
- What you can do after
- Evaluate when learned representations may replace manual feature engineering and when tabular data still requires deliberate feature work.
Outlier Detection in Python
Representation learning is the process of learning useful numerical representations of objects from data.
- How it is used
- Creates reusable numerical inputs when sufficient hand-labeled data is impractical.
- How deep it goes
- Explains embedding extraction and reuse, while stopping short of detailed network architecture or training mathematics.
- Examples
- Uses transformed images, image rotations, and tabulated mug-image embeddings as concrete examples.
- What you can do after
- Apply learned embeddings as inputs to downstream predictive models with less labeled data.
Read how Outlier Detection in Python explains representation learning →
Data Without Labels
Feature learning is the automated discovery of useful data representations or features by a model during training.
- How it is used
- Automates discovery of useful representations for feature detection, especially in image, text, audio, and video datasets.
- How deep it goes
- Defines supervised and unsupervised approaches, then details autoencoder components, training choices, applications, and limitations.
- Examples
- Uses autoencoders for dimensionality reduction, anomaly detection, data compression, and synthetic-data generation.
- What you can do after
- Apply it via neural networks or autoencoders to discover representations without manual feature engineering.
Read how Data Without Labels explains representation learning →
Related concepts
- deep learningused by · Deep learning is a machine-learning approach that uses neural networks with multiple computational layers to learn representations from data.
- neural networkuses · A neural network is a computational model of interconnected, parameterized units that learns transformations from inputs to outputs.
- data representationproduces · A data representation is a form or encoding in which information is structured for storage, processing, transmission, or interpretation.
- language modelincludes · A language model is a computational model that represents patterns in linguistic tokens to estimate sequence probabilities and predict or generate language from context.
- feature engineeringcontrasts with · Feature engineering is the creation, transformation, or selection of input features from data to provide useful representations for analysis or modeling.
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.