Multilabel classification

As explained in Deep Learning with R, Third Edition

Multilabel multiclass classification is a classification task in which each input may belong to multiple categories drawn from a set of more than two classes.

Multilabel classification assigns each input multiple labels from a shared set. Unlike single-label classification, an input can belong to several categories, and the number of labels may vary. section 4.0

Image tagging

An image containing both a cat and a dog receives both the “cat” and “dog” labels. Image classification systems can similarly tag every category present in an image. section 4.0 section 11.1

The distinction depends on whether categories are exclusive. Reuters wires assigned one topic illustrate single-label multiclass classification; allowing several topics would make the task multilabel multiclass classification. section 4.2

Figure 11.1 The three main computer vision tasks: classification, segmentation, and detection

Compare multilabel image classification with segmentation and object detection: classification assigns labels, segmentation partitions pixels, and detection adds bounding boxes. section 11.1

How Deep Learning with R, Third Edition covers it

How deep it goes

Defines multilabel classification and contrasts it with binary and single-label multiclass classification, without detailing model training.

Examples

Uses cat-and-dog image tagging, Reuters newswire topics, and Google Photos search across more than 20,000 image classes.

What it assumes

Builds from binary and multiclass classification definitions, introducing the multilabel distinction explicitly.

What you can do after

Evaluate/understand whether a classification problem permits multiple labels per input rather than mutually exclusive categories.

How it connects

Connects multilabel classification to binary classification, multiclass classification, image classification, image segmentation, and object detection.

liveBookmanning.com

Read Deep Learning with R, Third Edition on liveBook Get the book at manning.com

Also explained in

• 

  liveBookmanning.com
  Natural Language Processing in Action, Second EditionMulti-label classification assigns zero or more independent labels to each input example.
• 

  liveBookmanning.com
  Deep Learning with Python, Third EditionA classification task in which each example may belong to multiple classes at the same time, even though there are multiple possible classes overall.

Compare how 3 books explain multilabel classification →

Related concepts

• classifierkind of · A classifier is a predictive computational model that assigns inputs to one or more discrete classes or categories.
• single-label categorical classificationcontrasts with · A classification setup in which each input is assigned no more than one label from a set of classes.
• multi-hot tensoruses · A multi-hot tensor is a binary tensor in which multiple positions can be active simultaneously to represent selected labels or attributes.
• multiclass classificationkind of · Multiclass classification is a supervised prediction task that assigns each input to exactly one of more than two possible classes.
• classificationkind of · Classification is the task of assigning an input or observation to one or more categories from a defined set of labels.

This explainer was generated by Manning's knowledge graph from the text of Deep Learning with R, Third Edition. The section links point to the parts of the book it draws on.