Unlabeled data
Unlabeled data is a collection of examples that lack target labels or task-specific annotations.
Grokking Machine Learning
Data whose observations do not have associated target labels or outcome annotations.
- How it is used
- Supports unsupervised learning by providing observations from which algorithms extract patterns without target labels.
- How deep it goes
- Defines the absence of labels, notes that label status depends on the problem, and illustrates grouping without detailing algorithm internals.
- Examples
- Uses pet images without type tags and email records described by size and recipient count.
- What you can do after
- Understand how unlabeled observations can support similarity grouping and data simplification rather than label prediction.
Read how Grokking Machine Learning explains unlabeled data →
Outlier Detection in Python
Data whose instances do not have assigned target labels or annotations.
- How it is used
- Supports unsupervised outlier detection and complements supervised models when labeled and unlabeled records coexist.
- How deep it goes
- Explains distributional value, manual labeling, and classifier limitations without developing specific detection algorithms.
- Examples
- Uses detector-flagged records, near-threshold records, and a 100,000-row dataset with approximately 300 manually labeled examples.
- What you can do after
- Evaluate whether unlabeled records can support outlier detection or semisupervised modeling while recognizing labeling and assumption limits.
Read how Outlier Detection in Python explains unlabeled data →
Grokking Machine Learning, Second Edition
Data that does not include labels identifying target values or categories for the problem being considered.
- How it is used
- Frames unsupervised learning as a way to extract structure when target labels are unavailable.
- How deep it goes
- Defines unlabeled data with email and image examples, then names clustering, dimensionality reduction, and generative learning.
- Examples
- Uses email records without target columns and pet-image collections without animal or weight labels.
- What you can do after
- Evaluate/understand whether a dataset has a problem-specific target and identify suitable unsupervised uses.
Read how Grokking Machine Learning, Second Edition explains unlabeled data →
Natural Language Processing in Action, Second Edition
Unlabeled text is textual data that lacks human-provided category, entity, or other annotation labels.
- How it is used
- Supports unsupervised NLP systems that learn linguistic relationships and world knowledge from large text collections without manual labels.
- How deep it goes
- Explains the label-free learning principle through Word2Vec and word contexts, but stops short of detailing corpus preparation or training procedures.
- Examples
- Uses corpora mentioning Marie Curie, the Timbers, Portland, science, soccer, and cities to illustrate contextual learning.
- What you can do after
- Evaluate/understand how NLP models can extract relationships and knowledge from text without manually assigned target labels.
Read how Natural Language Processing in Action, Second Edition explains unlabeled data →
Building Specialized AI Systems
Unlabeled text is raw textual data without annotations or desired output labels.
- How it is used
- Uses raw domain language to adapt models before task-specific supervised fine-tuning.
- How deep it goes
- Explains the adaptation rationale and gives a masked-language-modeling implementation, stopping short of detailed training mechanics.
- Examples
- Uses financial news, clinical notes, research papers, technical documentation, and customer interactions as unlabeled corpora.
- What you can do after
- Apply it via domain-adaptive pretraining to align language priors with specialized vocabulary and writing patterns.
Read how Building Specialized AI Systems explains unlabeled data →
Building LLM Applications with DSPy
Data examples that lack target outputs or labels.
- How it is used
- Supports evaluating and optimizing summarization when examples lack reference summaries.
- How deep it goes
- Defines input-only examples and their use for evaluation and optimization, without detailing the optimization procedure.
- Examples
- Uses original documents without ground-truth summaries as summarization examples.
- What you can do after
- Apply it via DSPy to evaluate and optimize examples without ground-truth summaries.
Read how Building LLM Applications with DSPy explains unlabeled data →
Also covered in
- Build a Large Language Model (From Scratch)
- Introduction to Generative AI, Second Edition
- The Art of AI Product Development
- Human-in-the-Loop Machine Learning: Active learning and annotation for human-centered AI
- Data Without Labels
What to know first
- language modelA language model is a computational model that represents patterns in linguistic sequences to assign probabilities, predict, generate, or evaluate language.
- supervised learningSupervised learning is a machine-learning approach that learns predictive relationships from examples paired with known target outputs or labels.
- fine-tuningFine-tuning is the continued training of a pretrained model on targeted data or objectives to adapt its parameters or behavior to a task, domain, or use.
Related concepts
- datakind of · Data is recorded or represented information, observations, measurements, or values that can be stored, processed, communicated, or interpreted.
- unsupervised learningused by · Unsupervised learning is a machine-learning approach that discovers patterns, structure, or representations in data without labeled target outputs.
- active learningused by · Active learning is a machine-learning approach in which a learner selects informative data examples for labeling or feedback to improve a model.
- data annotationused by · Data annotation is the process of assigning labels, metadata, or structured judgments to data according to specified criteria.
- labeled datacontrasts with · Labeled data is a collection of examples paired with target values, categories, or other task-relevant annotations.
- random samplingused by · Random sampling is a method of selecting items from a population through chance rather than deterministic selection criteria.
Questions about unlabeled data
What is unlabeled data?
Unlabeled data is a collection of examples that lack target labels or task-specific annotations.
Which Manning books explain unlabeled data?
Grokking Machine Learning, Outlier Detection in Python, Grokking Machine Learning, Second Edition, Natural Language Processing in Action, Second Edition, Building Specialized AI Systems and Building LLM Applications with DSPy.
What should I know before learning unlabeled data?
Manning books that teach unlabeled data usually introduce language model, supervised learning, attention projection and fine-tuning first.
What is the difference between unlabeled data and labeled data?
Unlabeled data is a collection of examples that lack target labels or task-specific annotations. Labeled data is a collection of examples paired with target values, categories, or other task-relevant annotations.
What is unlabeled data a kind of?
Unlabeled data is a kind of data. Data is recorded or represented information, observations, measurements, or values that can be stored, processed, communicated, or interpreted.
What is unlabeled data also called?
Unlabeled data is also called unlabeled dataset.
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.





