A chi-squared divergence is a human-made measure of difference between two probability distributions P and Q with densities p and q relative to a common dominating measure, defined as D_χ²(P||Q) = ∫ (p(x) - q(x))² / q(x) dμ(x) = E_Q[(dP/dQ - 1)²]. It is the f-divergence generated by the convex function φ(t) = (t-1)², making it a member of the broader f-divergence family. The measure is always non-negative and equals zero iff P = Q almost everywhere. Unlike KL-divergence, it is not symmetric and may diverge when Q assigns near-zero mass where P has significant density. It persists through its use in goodness-of-fit tests, density ratio estimation, and as a computationally tractable divergence in variational inference. [formal: χ²-divergence | substrate: mind | horizon: a life | explicit: yes | epoch: 0.01]
Accepted ontology entry
chi-squared-divergence
A chi-squared divergence is a human-made measure of difference between two probability distributions P and Q with densities p and q relative to a common dominating measure, defined as D_χ²(P||Q) = ∫ (p(x) - q(x))² / q(x) dμ(x) = E_Q[(dP/dQ…
Definition
Why it is in scope
Chi-squared divergence is a human-made measure of difference between two probability distributions, defined as D_χ²(P||Q) = ∫ (p(x) - q(x))² / q(x) dx. It is the f-divergence generated by φ(t) = (t-1)² and serves as a computationally convenient alternative to KL-divergence in goodness-of-fit testing and density estimation. It persists through its role in mathematical statistics and hypothesis testing.
Names and aliases
- chi-squared-divergenceen · CANONICAL
Relations from this entry
- cmsl5bc2a06d5nobpsnqnh00cSERVES →
Chi-squared divergence is constructed as a computationally convenient divergence for comparing distributions, specifically used within goodness-of-fit testing and model comparison. It is built for the sake of evaluating how well a model fits observed data, i.e., it serves goodness-of-fit practice.
- f-divergenceINSTANCE_OF →
Chi-squared divergence is the f-divergence generated by φ(t)=(t-1)². It is a specific kind of f-divergence, with its convex function parameter fixing the divergence family.
- chi-squared-statisticDERIVED_FROM →
Which-came-first (Law 7): Pearson's 1900 statistic X² = Σ(O−E)²/E is the antecedent object. The chi-squared divergence is that same functional form, D_χ²(P||Q) = Σ(p−q)²/q, read as a measure on distributions — the statistic is the sample-size-scaled empirical instance (X² = n·D_χ²(empirical||model)) — and the general divergence reading was absorbed into the f-divergence framework (Csiszár 1966, φ(t)=(t−1)²). The statistic (1900) came first; the divergence construct is derived from it.
Relations to this entry
No accepted relations in this direction.
Record identity
- Created
- Sep 3, 2026, 11:26 PM UTC
- Content hash
- 6ec672f97058cb931609cc31d165aa7721986b85d9fd7648546afdab6a0350a9