The goal of this full-day workshop is to call on the research community to discuss the state-of-the-art and research challenges for supporting modeling and decision making under uncertainty in the computational and data sciences. The workshop will be held at IEEE VIS 2015 in Chicago. All registered attendees of VisWeek will be able to attend the workshop.

Motivation

The goals of this workshop are three-fold: a) to identify what methods and techniques have been successfully deployed into application domains, b) to understand when and why methods have failed or are not in use, c) to broaden the scope of work from mere ensemble analysis to support sensitivity analysis as well as the decision making process based on computational model.

Uncertainty is often talked about in numerous fields, however the meaning, definition, and usage highly varies between these fields. The relationship between uncertainty and parameter space analysis is also up to debate; it is clear a relationship exists, however the extent of this relationship is unclear. Uncertainty analysis (UA) can be thought of as an essential component of modeling and decision making. The underlying computational models have an inherent parameter space in which to explore and the ensembles are in fact comprised of several realizations of these models for varying parameters. Additional challenges lie in the sensitivity analysis (SA) of the models and both UA and SA are essential components to the decision making process. Other aspects of the decision making process include trading off multiple objectives as well as ranking of multiple possible solutions.

Goals

The goals of this workshop are three-fold: a) to identify what methods and techniques have been successfully deployed into application domains, b) to understand when and why methods have failed or are not in use, c) to broaden the scope of work from mere ensemble analysis to support sensitivity analysis as well as the decision making process based on computational model.

Uncertainty is often talked about in numerous fields, however the meaning, definition, and usage highly varies between these fields. The relationship between uncertainty and parameter space analysis is also up to debate; it is clear a relationship exists, however the extent of this relationship is unclear. Uncertainty analysis (UA) can be thought of as an essential component of modeling and decision making. The underlying computational models have an inherent parameter space in which to explore and the ensembles are in fact comprised of several realizations of these models for varying parameters. Additional challenges lie in the sensitivity analysis (SA) of the models and both UA and SA are essential components to the decision making process. Other aspects of the decision making process include trading off multiple objectives as well as ranking of multiple possible solutions.

To better understand the issues relating to uncertainty and parameter space, we must first survey what has been done, to date, to successfully handle the expression of uncertainty in practical applications. We believe the best way to assess this is to invite workshop participants to join in the conversation by presenting their successes and failures and bring in domain experts to help us understand how we, as visualization experts, can best address their needs.

The technical scope of this workshop will be determined by the accepted short-papers, however our invited domain speakers will ensure an alternative perspective to the problem. We will particularly target the inclusion of all visualization participants to encourage greater communication between the subfields. Specifically, work in decision making, cognition and perception, and model building will be requested since understanding in these areas is fundamental in progressing the state of the art.

Some of the questions we would like to ask include: