---
title: 'Credal Set Correspondence: Theory and Applications'
url: https://www.emergentmind.com/topics/credal-set-correspondence
type: topic
---

# Credal Set Correspondence: Theory and Applications

Credal set correspondence refers to formal relationships between sets of probability measures (credal sets) and other mathematical objects or representations—most notably, interval-valued probabilities, capacities, random set representations, convex analysis duals, or other imprecise-probability structures. Such correspondences enable transformation, extension, and analysis of imprecise probabilistic reasoning in a mathematically principled way. This article presents the key definitions, constructions, theoretical characterizations, and algorithmic consequences of credal set correspondence, as documented across measure-theoretic, convex-geometric, game-theoretic, and computational frameworks in recent research.

## 1. Credal Sets: Definitions and Foundational Notions

A credal set is any nonempty, convex, and typically closed (or compact) set of probability measures on a measurable space $(X, \Sigma)$ or, in the discrete case, on a finite set of outcomes or classes. Let $\mathcal{P}(X)$ denote the space of all probability measures on $(X, \Sigma)$. A credal set is then $K \subseteq \mathcal{P}(X)$, compact and convex in the weak topology [2604.09272].

For a finite outcome space $Y = \{y_1, \dots y_K\}$, a credal set $\mathcal{K}$ can be described as
\[
\mathcal{K} = \bigg\{ p \in \Delta^K : \underline{P}(y_k) \leq p_k \leq \overline{P}(y_k); \sum_{k=1}^K p_k = 1 \bigg\}
\]
where $\Delta^K$ is the $K$-simplex, and $[\underline{P}(y_k), \overline{P}(y_k)]$ are lower/upper probability bounds per outcome [2405.15047].

These sets support robust representations of epistemic uncertainty, encompassing not only precise Bayesian models $(K$ a singleton$)$ but full imprecise-probabilistic modeling as well.

## 2. Credal Set to Interval Domain Correspondence

A central correspondence is between credal sets and interval-valued probabilities (envelopes) for each event. Given a credal set $K$, for each measurable event $A \in \Sigma$, define the lower and upper probabilities:
\[
\underline{P}_K(A) = \inf_{\mu \in K} \mu(A), \qquad \overline{P}_K(A) = \sup_{\mu \in K} \mu(A)
\]
This maps $K$ to the interval $[\underline{P}_K(A), \overline{P}_K(A)] \subseteq [0,1]$.

Formally, for the space of credal sets $U(\mathcal{P}(X))$, define the envelope map $\varphi_A$:
\[
\varphi_A(K) = \bigl[\inf_{\mu \in K} \mu(A), \sup_{\mu \in K} \mu(A)\bigr]
\]
The mapping $\varphi_A: U(\mathcal{P}(X)) \to I_{[0,1]}$ is Scott-continuous under the reverse inclusion order on the interval domain $I_{[0,1]}$ [2604.09272].

This mapping extends to capacities (monotone set functions), producing lower and upper capacities that are monotone on $\Sigma$, and links to the theory of Choquet integration:
\[
\int f \, d\underline{P}_K = \inf_{\mu \in K} \int f \, d\mu
\]
for all non-negative measurable $f$, with analogous results for $\overline{P}_K$. Thus, the credal set–interval domain correspondence provides a complete and robust realisation of classical capacity theory and Choquet expected value calculations.

## 3. Convex, Dual, and Geometric Correspondences

Credal sets admit a variety of geometric and dual representations. In finite dimensions, the duality between credal sets and convex cones of almost desirable gambles $K$ (closed, convex cones in $\mathbb{R}^n$ containing the nonnegative orthant) is given by polarity [1705.09574]:
\[
K^0 = \{v \in \mathbb{R}^n : \langle v, x \rangle \ge 0 \; \forall x \in K\}
\]
and the intersection $K^0 \cap \mathcal{P}_n$ (the probability simplex) yields a credal set. Conversely, any credal set $C \subset \mathcal{P}_n$ yields the polar cone
\[
C^\perp = \{g \in \mathbb{R}^n : \langle p, g \rangle \ge 0\;\forall p \in C\}
\]
This polarity correspondence is a bijection, preserving conditioning and marginalization.

Extensions to lexicographic probabilities (full-rank stochastic matrices representing layers of preference) enable stricter, order-sensitive correspondences for strict desirability cones via lexicographic polarity [1705.09574]:
\[
D^{\circ} = \{A \in \mathbb{R}^{n \times n} : A g >_L 0_n\ \forall g \in D\}
\]
with the “$L$-credal set” $L(D) = D^{\circ} \cap T_{n,n}$ (full-rank stochastic matrices), and their mutual bijection with strict desirability cones.

## 4. Algorithmic and Computational Correspondences

Credal set correspondences support efficient algorithms for probabilistic inference under imprecise information:

- **Credal Sentential Decision Diagrams (CSDDs):** Replace precise local probabilities by local credal sets $[a_i^{(L)}, a_i^{(U)}]$ at disjunctive nodes, inducing global credal sets as convex hulls of PSDDs consistent with these constraints. Inference is performed by bottom-up traversal with local LPs, yielding lower and upper marginal (and conditional) probabilities as extrema over the global credal set [2008.08524].
- **Decision-making under ambiguous evidence:** For Bayesian networks, uncertain evidence is encoded as credal virtual or soft evidence; network updating proceeds by augmenting with auxiliary credal nodes, and the computed posterior marginals exactly correspond to the lower and upper bounds obtained via all consistent Bayesian updates [1802.05639].
- **Model averaging in deep learning:** The credal wrapper transforms $M$ predictive distributions from Bayesian neural networks or deep ensembles into a polytope credal set $\mathcal{K}$ and yields a unique prediction via the intersection-probability transform, which exactly selects the distribution in $\mathcal{K}$ that is an affine combination of the lower and upper bounds with a data-dependent normalization coefficient [2405.15047].
- **Learning from sets:** In credal learning theory, generalization and adaptation bounds for set-valued uncertainties are expressed as the supremum over all $P \

Source: https://www.emergentmind.com/topics/credal-set-correspondence