BEGIN:VCALENDAR VERSION:2.0 PRODID:-//Talks.cam//talks.cam.ac.uk// X-WR-CALNAME:Talks.cam BEGIN:VEVENT SUMMARY:Decision-making under uncertainty: Using subjective probabilistic judgements for decision support in pollinator abundance and food security - Martine Barons (University of Warwick) DTSTART:20190704T111000Z DTEND:20190704T113000Z UID:TALK126808@talks.cam.ac.uk CONTACT:INI IT DESCRIPTION:Hunger and food poverty is on the increase even in developed n ations like the UK\, USA\, Canada &\; Australia. With a growing popula tion and a finite planet\, there is urgent need for action\, but in such a large\, complex system identifying the most effective action requires dec ision support.
Food security exists when all people\, at all times\, have physical and economic access to sufficient\, safe and nutritious foo d to meet their dietary needs and food preferences for an active and healt hy life.
In order to provide decision support\, it is necessary to e licit probability distributions to evaluate subjective expected utility sc ores associated with ameliorating policies that might be enacted. When the underlying process model is extremely large and complex\, this brings its own peculiar challenges. It is first necessary to elicit the overall\, ag reed structure describing in broad terms the underlying nature of the syst em from representatives of all domain experts across the system as a whole . We have now shown that this can be done formally and consistently with p robability models if the elicitations concern the elicitation of dependenc es &ndash\; formally termed irrelevances (Smith\, Barons and Leonelli (201 6)). Within a probability model\, these irrelevance statements then transf orm into assertions about various conditional independence statements. The se\, in turn\, can be used to determine how the system can be divided up i nto (conditionally) independent segments. The quantitative expert judgeme nts associated with each segment of the process can then be delegated to a relevant panel of experts. The implicit (albeit virtual) owner of beliefs expressed in the system will be referred to as the supraBayesian \, mean ing that the decision-making group acts as a single person would and it is her coherence that we are concerned with. Under suitable conditions it ca n then be shown that the elicited overarching structure can compose these judgments together to form a coherent probabilistic model to score differe nt options available to the user\, termed an integrating decisions support system (IDSS).

One element of the overarching food poverty mode ls is food supply\, and key to parts of this is an abundant and healthy po pulation of pollinating insects to pollination services for food. In 2014 the UK government undertook a consultation and produced their pollinator s trategy for the next 10 years &ldquo\;to see pollinators thrive\, providin g essential pollination services and benefits for food production\, the wi der environment and everyone.&rdquo\; However\, the evidence base on the complex system driving pollinator vigour and numbers is patchy and held in disparate domains of expertise\, making the evaluation of policy options problematic. In this talk I will describe how we are in the process of de veloping an IDSS based on these theoretical developments\, and how a proba bilistic model for pollinator abundance incorporating structured expert el icitation will then form a sub-module of this IDSS for policies relating t o household food insecurity.

J. Q. Smith\, M.J. Barons\, and M. Le onelli. Coherent inference for integrating decision support systems. arXiv \, 2016. http://arxiv.org/abs/1507.07394.

Co-authors: Jim Q. Smi th\, Manuele Leonelli
LOCATION:Seminar Room 1\, Newton Institute END:VEVENT END:VCALENDAR