DeepAI AI Chat
Log In Sign Up

A Theory of Uncertainty Variables for State Estimation and Inference

by   Rajat Talak, et al.

Probability theory forms an overarching framework for modeling uncertainty, and by extension, also in designing state estimation and inference algorithms. While it provides a good foundation to system modeling, analysis, and an understanding of the real world, its application to algorithm design suffers from computational intractability. In this work, we develop a new framework of uncertainty variables to model uncertainty. A simple uncertainty variable is characterized by an uncertainty set, in which its realization is bound to lie, while the conditional uncertainty is characterized by a set map, from a given realization of a variable to a set of possible realizations of another variable. We prove Bayes' law and the law of total probability equivalents for uncertainty variables. We define a notion of independence, conditional independence, and pairwise independence for a collection of uncertainty variables, and show that this new notion of independence preserves the properties of independence defined over random variables. We then develop a graphical model, namely Bayesian uncertainty network, a Bayesian network equivalent defined over a collection of uncertainty variables, and show that all the natural conditional independence properties, expected out of a Bayesian network, hold for the Bayesian uncertainty network. We also define the notion of point estimate, and show its relation with the maximum a posteriori estimate.


page 1

page 8


Bayesian Decision Theory and Stochastic Independence

Stochastic independence has a complex status in probability theory. It i...

Tractability through Exchangeability: A New Perspective on Efficient Probabilistic Inference

Exchangeability is a central notion in statistics and probability theory...

A Rational Distributed Process-level Account of Independence Judgment

It is inconceivable how chaotic the world would look to humans, faced wi...

Linearity Properties of Bayes Nets with Binary Variables

It is "well known" that in linear models: (1) testable constraints on th...

A Representation of Uncertainty to Aid Insight into Decision Models

Many real world models can be characterized as weak, meaning that there ...

Expected Utility Networks

We introduce a new class of graphical representations, expected utility ...

Relevant Explanations: Allowing Disjunctive Assignments

Relevance-based explanation is a scheme in which partial assignments to ...