BioMaths Colloquium Series - 2018/19
12 April 2019 - 3pm Robert Recorde Room
(Computational Foundry, Bay Campus)
Applications of Hidden Markov Models in Ecology
Prof Byron Morgan
(School of Statistics, University of Kent)
Our BioMaths Colloquium Series resumes for the spring term with a seminar by Prof Byron Morgan, from the School of Statistics, University of Kent. Byron is an Emeritus Professor of Statistics and Co-Director of the National Centre for Statistical Ecology (NCSE). He is broadly interested in Applied Statistics, biometry, and statistical ecology, ranging from stochastic models for molecular biology, parameter redundancy in ecological models, to applications of Bayesian methods and population dynamics.
Abstract
Abstract
Several standard models in common use in Statistical Ecology can be formulated in terms of hidden Markov Models (Zucchini et al, 2016). In this talk we summarise the essentials of hidden Markov modelling, including the forward algorithm, which provides the unifying structure for likelihood construction, and for efficient likelihood optimisation. Different applications require particular modifications, and this is shown through a range of illustrations, involving batch-marked animals in capture recapture, dynamic stochastic models for seasonal insect data, integrated population modelling, modelling population survey data and multi-species indicators.
References:
Cowan , L., Besbeas, P. T., Morgan, B.J.T. and Schwarz, C. (2017) Hidden Markov models for extended batch data, Biometrics, 73, 1321-1331.
Besbeas, P. T. and Morgan, B.J.T. (2018) A general framework for modelling population survey data, in revision.
Besbeas, P. T. and Morgan, B.J.T. (2019) Exact inference for integrated population modeling, Biometrics, in press.
Zucchini, W., MacDonald, I. L. and Longrock, R. (2016) Hidden Markov Models for Time Series: An Introduction using R, Second Edition, Chapman & Hall, CRC press, Boca Raton.
Cowan , L., Besbeas, P. T., Morgan, B.J.T. and Schwarz, C. (2017) Hidden Markov models for extended batch data, Biometrics, 73, 1321-1331.
Besbeas, P. T. and Morgan, B.J.T. (2018) A general framework for modelling population survey data, in revision.
Besbeas, P. T. and Morgan, B.J.T. (2019) Exact inference for integrated population modeling, Biometrics, in press.
Zucchini, W., MacDonald, I. L. and Longrock, R. (2016) Hidden Markov Models for Time Series: An Introduction using R, Second Edition, Chapman & Hall, CRC press, Boca Raton.
The discussions will continue over tea and coffee after the seminar.
No comments:
Post a Comment