Decision Support

Definition of Decision Support

Decision support provides advice in the presence of uncertainty, particularly where a decision needs to be made with incomplete data or knowledge. The known status of multiple influencing factors in the decision-making process is used to estimate the likelihood of an outcome of interest. Decision support can be provided using various approaches for medical diagnosis, longitudinal monitoring, technology evaluation and risk assessment. Various sources of information can be used to probabilistically quantify the causal relationships in the decision-making process. These include the literature, observational or experimental data and expert knowledge.

Who works in Decision Support

Paul Wu
Jim Lewis
Sandra Johnson
Cathy Hargrave

What Expertise do we have?

Workshops for developing conceptual models for decision support

CART analysis

Classification and regression tree analysis can be used to determine the key factors influencing a particular outcome of interest.

Bayesian meta-analysis

Pooling of multiple studies to inform an evidence-based approach. Bayesian meta-analysis better able to handle between study variations

Expert elicitation techniques – Link to Expert Elicitation document

Where insufficient data or knowledge exists, experts in a field can provide valuable information. Elicitation techniques can be used to:

• Ask experts to define important factors influencing specific outcomes of interest in order to define a model of the decision making process,
• Elicitation techniques where experts can also provide a degree of belief about specific statements relating to the decision-making process
• Obtain consensus for the above by combining the opinion of multiple experts

Data modelling for decision-support

Spatio-temporal models and hierarchical models

Bayesian networks

Link to BN document (coming soon)

Case studies: Concrete examples of what we have done

Airports of the future
Peak energy usage
Image-guided radiotherapy
Sandra’s work
AIMs (Paul and Julie)

Links to other capabilities: (e.g., Design / Optimisation)

Bayesian Networks
Expert elicitation
Modelling – spatio-temporal, hierarchical models

Contact

Sandra Johnson – sandra.johnson@qut.edu.au
Jim Lewis – james.lewis@connect.qut.edu.au
Cathy Hargrave – c.hargrave@qut.edu.au
Denise Beaudiquin – d.beaudequin@qut.edu.au