黑料福利网

Topic 

Distortion models for estimating human error probabilities

Abstract

Human Reliability Analysis aims at identifying, quantifying and proposing solutions to human factors causing hazardous consequences. Quantifying the influence of the human factors gives rise to human error probabilities, whose estimation is a cumbersome problem. Since these human factors are usually related to other organisational or technological factors, it has been proposed to apply probabilistic
graphical models, such as Bayesian or credal networks.

However, these can be problematic when conditional probabilities on missing data are involved. While the solutions proposed so far combine frequentist and subjective approaches and are in general not robust to small modifications in the dataset, in this talk I propose an alternative based on distortion models, which are a type of imprecise probabilities.

I will review some of the main distortion models that have been considered in the literature, provide a comparison in terms of a number of axiomatic properties, and show an application of some of them in the context of human reliability, showing that the proposal is consistent with the previous studies while giving rise to robust estimations.

About the speaker

Enrique Miranda is full Professor at the Department of Statistics and Operations Research at the University of Oviedo (Spain). Between 2003 and 2008 he was assistant professor at Rey Juan Carlos University in Madrid (Spain).

His research has focused on different aspects of imprecise probability theory. Imprecise probabilities arise in situations where the existence of imprecise or vague information makes it unadvisable to work with a precise and unique probability
distribution. In those situations, it may be useful (and more reasonable) to consider alternative models, such as sets of probability measures, upper and lower probabilities, Choquet capacities, etc.

In his work, he has studied some of the theoretical challenges that arise with these models, and more specifically investigated coherent lower previsions, sets of desirable gambles, random sets, and non-additive measures.

In addition, he has served at the Executive Committee of the Society for Imprecise Probabilities: Theories and Publications (SIPTA), of which he is currently President, and he is at the editorial board of the International Journal of Approximate Reasoning.

Your host

Cassio de Campos, Full Professor at the department of Mathematics and Computer Science.

Registration is required but free of charge.