← Back to catalogue
Published

probability

vr.tr.probability · thing-q9492

Let an agent handle probability by definition, interpretation, calculation and communication of risk, avoiding common fallacies.

Thing Registry Cross-cutting context XCT.QLT
Bundle → Layer → Finding → QuestionsClassifiersWhat it isWhy it existsDistinguishing featuresWhat robots and AI may and may not doMoral aspectsOwnersLinks to other meta-modelsIdentity and identifiersDirect propertiesRecognitionCapabilities and actionsHazards and failure modesStandards and interfacesContext of useSources

Bundle → Layer → Finding → Questions Filled

4 bundles · 8 layers · 8 findings · 16 questions

Calculation Computing.

Event

Probability of event.

Event

Event probability.

  1. What is the probability of this event? measurement
  2. Which assumptions are used? boundary

Conditional

Bayes.

Conditional

Conditional probability.

  1. What is the probability given this evidence? measurement
  2. Is the base rate included? boundary
Interpretation Meaning.

Meaning

Frequentist or Bayesian.

Meaning

Interpretation.

  1. Is the probability a long-run frequency or a degree of belief? definition
  2. Does it matter here? boundary

Model

Assumptions.

Model

Model assumptions.

  1. Which model generates the probability? provenance
  2. How sensitive is it to assumptions? boundary
Communication Risk.

Format

Natural frequencies.

Format

Risk format.

  1. Would natural frequencies such as 1 in 100 be clearer? action
  2. Is absolute risk given alongside relative risk? boundary

Fallacies

Errors.

Fallacies

Fallacies.

  1. Is a common fallacy such as the gambler fallacy involved? boundary
  2. How can it be corrected? action
Applications Uses.

Medical

Tests.

Medical

Medical tests.

  1. What is the probability of disease given a positive test, in general terms? measurement
  2. Should results be discussed with a clinician? action

Gambling

Odds.

Gambling

Gambling odds.

  1. What are the true odds compared with offered odds? measurement
  2. Where can someone get help with problem gambling? action

Classifiers Filled

Family
Thing Registry
Category
Cross-cutting context
Entry kind
thing
Plane
XCT
Domain
XCT.QLT
Other names and narrower kinds
availability, loss probability, error probability, joint probability, topic probability, geometric probability, probability of acceptance, negative probability, confidence level, tail probability, Probability of precipitation, sticking probability

What it is Filled

A numerical measure between 0 and 1 of how likely an event is, formalised by Kolmogorov axioms and interpreted in frequentist, Bayesian and other ways; related notions include joint, conditional and error probabilities.

Why it exists Filled

Let an agent handle probability by definition, interpretation, calculation and communication of risk, avoiding common fallacies.

Distinguishing features Filled

  • Measure of likelihood
  • Axiomatic foundation
  • Several interpretations
  • Easily misunderstood

What robots and AI may and may not do Filled

Must not

  • Present a probability as a certainty or a rare event as impossible.
  • Use relative risk alone to exaggerate a benefit or harm.
  • Misstate the probability of a person's guilt from match evidence.
  • Help design games or bets that mislead players about their odds.

Only with a human decision

  • Using probabilistic scores in decisions about a person, such as credit, bail or diagnosis.

May

  • Compute and explain probabilities, stating assumptions and the interpretation used.
  • Communicate risk with absolute and relative figures side by side.

Moral aspects Filled

  • Misread probabilities have led to wrongful convictions and poor medical choices.
  • Risk scores applied to people can encode bias.

Who is affected

  • People judged by risk scores
  • Patients and juries

Owners Filled

Steward

Nobody: a mathematical concept held in common.

Links to other meta-models Filled

parent

  • Q126818 - registry parent class
  • Q21170479 - registry parent class
  • Q246672 - registry parent class

related

  • mathematical object and dimensionless quantity - category
  • probability theory and statistics - fields
  • odds and risk - expressions
  • decision making - application

What else AI and robots need to interact with it Filled

Identity and identifiers required Filled

  • Vercy registry: vr.tr.probability
  • Wikidata: Q9492 (https://www.wikidata.org/wiki/Q9492)

Direct properties not applicable Not applicable

Not applicable

Plane XCT: no invented physical properties.

Recognition optional Filled

  • Values from 0 to 1 or 0 to 100 percent
  • Notation P(A)
  • Odds are a related but different expression
  • Not physical; numbers, fractions and percentages.

Capabilities and actions required Filled

  • compute probabilities of events
  • apply conditional probability and Bayes theorem
  • communicate risk clearly
  • spot common fallacies

Hazards and failure modes required Filled

  • Base rate neglect
  • Gambler fallacy
  • Confusing relative and absolute risk

Standards and interfaces required Filled

  • ISO 3534 statistics vocabulary

Context of use required Filled

  • Reasoning under uncertainty in science, engineering, medicine, finance and daily life.
  • classical probability
  • frequentist probability
  • Bayesian probability
  • conditional and joint probability
  • geometric probability

Sources Filled

  1. Wikidata item Q9492: probability - identity and sense of the item
  2. Wikipedia: Probability - general description of the item

Open questions

  • Should interpretations be separate entries?
  • How should risk communication guidance be linked?
  • How should odds be related?

Machine files

Card as JSON Specification Card backlog vercy.card/1

Provenance

thing registry research (pass 2) · unreviewed

Built from: models/things/publications/thing-q9492/spec.json