Model-robust Bayesian design through Generalised Additive Models for monitoring submerged shoals

08/30/2022
by   Dilishiya De Silva, et al.
0

Optimal sampling strategies are critical for surveys of deeper coral reef and shoal systems, due to the significant cost of accessing and field sampling these remote and poorly understood ecosystems. Additionally, well-established standard diver-based sampling techniques used in shallow reef systems cannot be deployed because of water depth. Here we develop a Bayesian design strategy to optimise sampling for a shoal deep reef system using three years of pilot data. Bayesian designs are generally found by maximising the expectation of a utility function with respect to the joint distribution of the parameters and the response conditional on an assumed statistical model. Unfortunately, specifying such a model a priori is difficult as knowledge of the data generating process is typically incomplete. To address this, we present an approach to find Bayesian designs that are robust to unknown model uncertainty. This is achieved through couching the specified model within a Generalised Additive Modelling framework and formulating prior information that allows the additive component to capture discrepancies between what is assumed and the underlying data generating process. The motivation for this is to enable Bayesian designs to be found under epistemic model uncertainty; a highly desirable property of Bayesian designs. Our approach is demonstrated initially on an exemplar design problem where a theoretic result is derived and used to explore the properties of optimal designs. We then apply our approach to design future monitoring of sub-merged shoals off the north-west coast of Australia with the aim of significantly improving on current monitoring designs.

READ FULL TEXT

page 3

page 13

page 19

page 31

research
04/29/2020

Optimal Designs for Kriging Models with Multiple Responses

Exact optimal designs for efficient prediction in simple and ordinary bi...
research
08/28/2018

Bayesian model-based spatiotemporal survey design for log-Gaussian Cox process

In geostatistics, the design for data collection is central for accurate...
research
03/20/2020

Sequential Bayesian Experimental Design for Implicit Models via Mutual Information

Bayesian experimental design (BED) is a framework that uses statistical ...
research
06/10/2022

Bayesian Design with Sampling Windows for Complex Spatial Processes

Optimal design facilitates intelligent data collection. In this paper, w...
research
04/02/2019

Optimal designs for model averaging in non-nested models

In this paper we construct optimal designs for frequentist model averagi...
research
11/15/2021

Where to Drill Next? A Dual-Weighted Approach to Adaptive Optimal Design of Groundwater Surveys

We present a novel approach to adaptive optimal design of groundwater su...
research
03/28/2023

A reinforced learning approach to optimal design under model uncertainty

Optimal designs are usually model-dependent and likely to be sub-optimal...

Please sign up or login with your details

Forgot password? Click here to reset