There is a significant need for principled uncertainty reasoning in machine learning systems as they are increasingly deployed safety-critical domains. A new approach with uncertainty-aware regression-based neural networks (NNs), based on evidential distributions aleatoric and epistemic uncertainties, shows promise over traditional deterministic methods typical Bayesian NNs, notably the capabil...

In applications of domain adaptation, there may exist multiple source domains, which can provide more or less complementary knowledge for pattern classification in the target domain. order to improve accuracy, a decision-level combination method is proposed multisource adaptation based on evidential reasoning. The results obtained from different domains usually have reliabilities/weights, are c...

“The Evidential Foundations of Probabilistic Reasoning” by David A. Schum “. . . contains a collection of thoughts . . . ” (p. 1) on issues related to evidence and to inference tasks based on evidence. The study of such issues is best summarized by an expression introduced in chapter 1: “Science of Evidence.” The Science of Evidence tries “. . . to treat the study of evidence as having a life o...

In this paper, we look at reasoning with evidence and facts in criminal cases. We show how this reasoning may be analysed in a dialectical way by means of critical questions that point to typical sources of doubt. We discuss critical questions about the evidential arguments adduced, about the narrative accounts of the facts considered, and about the way in which the arguments and narratives are...

Isaac is a rule-based visual language for mobile robots using evidential reasoning and a fuzzy inference engine. A prototype inference engine for Isaac has been implemented, permitting experiments with the Isaac language. This paper discusses this inference engine, describes some preliminary experiences with programming Isaac rulesets, and proposes future optimizations and enhancements to the i...