Quantitative methods: Bayesian inference, Bayesian thinking

Abstract

Bayesian methods are genuine inferences which admit genuine confidence intervals and more truly reflect the manner in which knowledge, especially statistical knowledge, is obtained about phenomena and the underlying processes which are assumed to govern them.

Mapping Manila's Mega-Urban Region

“…population living in 'slums'). Other spatial statistical approaches, such as geographically weighted regression (Fotheringham, Brunsdon, & Charlton, 2000), Bayesian methods (Withers, 2002) and agent-based modelling (Xie et al, 2007) provide alternative ways of analysing population-related issues in MURs.…”

“…Kwan, 2004;Matthews, Detwiler, & Burton, 2005;Withers, 2002). This paper builds on earlier efforts by Jones (2002Jones ( , 2004Jones ( , 2005 in analysing demographic trends and patterns in selected Southeast Asian MURs.…”

A GIS-based Bayesian approach for analyzing spatial–temporal patterns of intra-city motor vehicle crashes

“…In the public health area, application of Bayesian methods in disease mapping, risk assessment and prediction are numerous (Besag and Newell, 1991;Wakefield and Morris, 2001;Wakefield et al, 2000;Waller et al, 1997). The ability to incorporate prior knowledge without the restriction of classical distributional assumptions makes Bayesian inference a potent forecasting tool in a wide variety of fields (Withers, 2002). The Bayesian approaches, from empirical Bayes to full Bayes, were also implemented in some crash analysis studies to estimate crash risk and predict crash frequency (Brü de and Larsson, 1988;Hauer, 1992Hauer, , 2002Mountain et al, 1996).…”

“…One of the major advantages of the Bayesian approach is its ability to forecast risks accurately even in the presence of sparse data or rare events (Withers, 2002). In the public health area, application of Bayesian methods in disease mapping, risk assessment and prediction are numerous (Besag and Newell, 1991;Wakefield and Morris, 2001;Wakefield et al, 2000;Waller et al, 1997).…”

Modelling urban growth: Integrating both physical and human dimensions in developing world context

“…Various dynamic modelling approaches like System Dynamics (SD) (Guo et al, 2001;Mohapatra et al, 1994), Bayesianbased model (e.g. Bayesian-neural, Bayesian-belief network) (Withers, 2002), or artificial intelligence (artificial-neural-network, support-vector-machine) are available in the scientific literature. The implementation of these modelling approaches in spatial studies is nevertheless inadequate because they have no structure to represent the explicit spatial dimensions of development actors (e.g.…”