Selection bias

As explained in Deep Learning with R, Third Edition

Sampling bias is systematic distortion in measurements caused by a data-collection process that selects examples in relation to the quantity being predicted.

Selection bias, also called sampling bias, occurs when the data-collection process interacts with the quantity being predicted, producing systematically distorted measurements. A model can only learn patterns present in its training data, so predictions become unreliable when collected examples do not represent the population or future cases. section 6.1.2

The 1948 election poll

A phone survey predicted Dewey would win the US presidential election, but Truman won. Phone users were not a random sample: they were more likely to be richer, conservative, and to support Dewey. The survey therefore measured a selected population rather than the voting population. section 6.1.2

“DEWEY DEFEATS TRUMAN”: a famous example of sampling bias

Use the figure to inspect the canonical “DEWEY DEFEATS TRUMAN” example of sampling bias. section 6.1.2

How Deep Learning with R, Third Edition covers it

How deep it goes

Defines the distortion, explains its mechanism, and illustrates it historically, but stops short of detailing correction methods.

Examples

Uses the 1948 US presidential election and its unrepresentative phone survey as the central example.

What it assumes

Assumes basic ideas about data collection and prediction while explaining sampling bias from its definition and example.

What you can do after

Diagnose how a data-collection process can distort measurements and recognize when a sample is not representative.

How it connects

Connects sampling bias with data collection, nonrepresentative data, training data, prediction, and concept drift.

liveBookmanning.com

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

Also explained in

• 

  liveBookmanning.com
  Deep Learning with Python, Third EditionA bias introduced when the data collection process produces a sample that is not representative of the target population.
• 

  liveBookmanning.com
  Grokking StatisticsSelection bias is systematic distortion caused by including some groups in a sample more readily than others, making those groups overrepresented.
• 

  liveBookmanning.com
  Experimentation for EngineersSystematic measurement error caused by collecting samples under conditions that differ across the populations or variants being compared.

Compare how 4 books explain selection bias →

Related concepts

• biaskind of · Bias is a systematic deviation or tendency in data, measurements, estimates, judgments, or outputs that skews them relative to a relevant target or impartial baseline.
• samplingproduced by · Sampling is a procedure for selecting or generating observations, values, or outcomes from a larger population, signal, or probability distribution according to a specified method.
• randomizationcontrasts with · Randomization is the deliberate use of chance to select, assign, arrange, or vary elements of a process.
• calibration datacontrasts with · Calibration data is data used to measure, estimate, or adjust a model’s parameters, outputs, or behavior for a specified calibration objective.
• social mediaproduced by · Social media is a class of online services that enables users to create, share, discover, and interact with content and with one another.

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.