[MARMAM] Developing Bayesian Belief Networks (BBNs) to improve decision-making during mass stranding events

Stockin, Karen K.A.Stockin at massey.ac.nz
Thu Sep 7 15:50:18 PDT 2017


Dear colleagues
Just a reminder re the upcoming full day workshop entitled " Developing Bayesian Belief Networks (BBNs) to improve decision-making during mass stranding events” at the upcoming 22nd Biennial Conference of the Marine Mammal Society. The workshop will take place on Sunday 29th October from 0800 to 1700h and assemble scientists, vets, NGOs, government agencies and other stakeholders whom have experience and/or vested interest in (mass) stranding events. The workshop will engage expert opinion on the different parameters that can affect the likelihood of survivorship of refloated individuals and work towards establishment of predicted probabilities that can affect the outcome of rescue attempts. The overarching goal will be to prepare a submission for publication that addresses the application of BBNs to assess probability of survivorship in refloated whales.

Background

Despite costly, and often logistically challenging attempts to rescue live whales, there is often a lack of scientific evaluation underpinning current decision-making processes. Notably, matters of conservation (survivorship/fitness) and animal welfare (impacts of refloatation), remain largely undetermined. Historically, animal welfare science and conservation have been regarded as separate disciplines, with dissimilar objectives that often conflict. However, the newly emerging field of conservation welfare integrates synergies between two scientific disciplines with the aim of improving outcomes for both the species (conservation) and individual animals (welfare). This workshop addresses the issue of Artificial Intelligence (AI) in the form of Bayesian Belief Networks (BBN) to challenge the human perceptions and psychology associated with whale mass strandings, while incorporating scientific evaluation into decision-making processes. AI tools are becoming increasingly popular to address an array of complex environmental problems, but have yet to be applied effectively at the interface between science and public interaction. The overarching goal is to apply recent technological innovations to an age-old problem, in order to provide a vital nexus between conservation and animal welfare sciences.

Workshop Summary
Decision-making processes required by authorities during live stranding events are typically fraught with difficulties due to complicated, often interlinked variables, including but not limited to logistics, ethics, public perceptions and animal welfare. Bayesian Belief Networks (BBNs) are a graphical rule based modelling technique that have recently emerged as a useful research and management tool. BBNs can provide a visual depiction of the causal linkages between multiple environmental drivers and ecological state. Notably, in the absence of empirical data, BBNs can be constructed solely upon expert opinion, with subsequent independent assessment applied to assess the prediction accuracy of the model. This workshop aims to convene and engage individuals with relevant live stranding event experience to determine as a collective, key parameters and their predicted probability of influence on survivorship of refloated cetacea post-stranding event.

Registration for the workshop can be completed via the conference website (https://www.xcdsystem.com/smm/member/index.cfm)

We look forward to seeing you in Halifax!

Karen A Stockin, PhD
Director, Coastal-Marine Research Group

Institute of Natural and Mathematical Sciences
Massey University
Private Bag 102 904
North Shore
Auckland 0745

Tel: 09 4140800 ext 43614
Tel: 09 2136614 (direct dial)
Mobile: 021 423 997
Email: k.a.stockin at massey.ac.nz<mailto:k.a.stockin at massey.ac.nz>
Web: http://cmrg.massey.ac.nz<http://cmrg.massey.ac.nz/>

https://www.massey.ac.nz/massey/learning/colleges/college-of-sciences/staff-list.cfm?stref=926050

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.uvic.ca/pipermail/marmam/attachments/20170907/f1be69e5/attachment.html>


More information about the MARMAM mailing list