liveBook · Manning

Bias

As explained in Deep Learning with R, Third Edition

Bias is a systematic tendency in data, judgments, or model outputs that produces uneven or distorted representations of people or phenomena.

Bias is a systematic tendency in data, judgments, or model outputs that creates uneven or distorted representations. Machine-learning models can encode the biases of the people who label their training data. section 6.1.1

Trustworthiness from a face

A system that rates a person’s trustworthiness from a photograph may reproduce annotators’ prejudices. Its algorithmic output can make subjective judgments appear objective and operationalize them at scale, affecting real people. section 6.1.1

Bias becomes especially consequential when:

  • Labels reflect human prejudices rather than a valid property being measured. section 6.1.1
  • A model gives biased judgments a veneer of legitimacy and applies them broadly. section 6.1.1

How Deep Learning with R, Third Edition covers it

How deep it goes
Defines how labeling bias becomes encoded in models and scaled, but stops short of presenting mitigation methods.
Examples
Uses facial trustworthiness ratings to illustrate prejudiced labels, questionable targets, and scaled algorithmic harm.
What you can do after
Evaluate machine-learning proposals for questionable targets, biased labels, and harmful effects on people.
How it connects
Connects bias to human judgment, data labeling, machine-learning models, and algorithmic legitimacy.
Cover of Deep Learning with R, Third Edition
liveBookmanning.com
Read Deep Learning with R, Third Edition on liveBook Get the book at manning.com

Also explained in

Compare how 28 books explain bias →

Related concepts

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.