A Bayesian Nonparametric Approach for Time Series Clustering

Nieto‐Barajas, Luis E.; Contreras‐Cristán, Alberto

doi:10.1214/13-ba852

Cited by 37 publications

(13 citation statements)

References 29 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Several Bayesian nonparametric studies have been specially targeted to clustering temporally evolving phenomena. For example, in a recent work Nieto-Barajas and Contreras-Cristán (2014) accommodated the temporal effects in time series data using a first order autoregressive process. In our model, we used a simple and feasible solution given by introducing natural cubic splines that correct for temporally dependent confounding effects, adjusting for seasonal and long-term trends and weather variables such as temperature.…”

Section: Discussionmentioning

confidence: 99%

Analysing the health effects of simultaneous exposure to physical and chemical properties of airborne particles

Pirani

Best

Blangiardo

et al. 2015

Environment International

View full text Add to dashboard Cite

BackgroundAirborne particles are a complex mix of organic and inorganic compounds, with a range of physical and chemical properties. Estimation of how simultaneous exposure to air particles affects the risk of adverse health response represents a challenge for scientific research and air quality management. In this paper, we present a Bayesian approach that can tackle this problem within the framework of time series analysis.MethodsWe used Dirichlet process mixture models to cluster time points with similar multipollutant and response profiles, while adjusting for seasonal cycles, trends and temporal components. Inference was carried out via Markov Chain Monte Carlo methods. We illustrated our approach using daily data of a range of particle metrics and respiratory mortality for London (UK) 2002–2005. To better quantify the average health impact of these particles, we measured the same set of metrics in 2012, and we computed and compared the posterior predictive distributions of mortality under the exposure scenario in 2012 vs 2005.ResultsThe model resulted in a partition of the days into three clusters. We found a relative risk of 1.02 (95% credible intervals (CI): 1.00, 1.04) for respiratory mortality associated with days characterised by high posterior estimates of non-primary particles, especially nitrate and sulphate. We found a consistent reduction in the airborne particles in 2012 vs 2005 and the analysis of the posterior predictive distributions of respiratory mortality suggested an average annual decrease of − 3.5% (95% CI: − 0.12%, − 5.74%).ConclusionsWe proposed an effective approach that enabled the better understanding of hidden structures in multipollutant health effects within time series analysis. It allowed the identification of exposure metrics associated with respiratory mortality and provided a tool to assess the changes in health effects from various policies to control the ambient particle matter mixtures.

show abstract

Section: Discussionmentioning

confidence: 99%

Analysing the health effects of simultaneous exposure to physical and chemical properties of airborne particles

Pirani

Best

Blangiardo

et al. 2015

Environment International

View full text Add to dashboard Cite

show abstract

“…When convergence is attained, to avoid label switching ( Stephens, 2000 ), we follow Nieto-Barajas and Contreras-Cristan (2014) and Dahal (2006) and choose the representative clustering structure which minimises the deviation from a pairwise clustering matrix which, taking into account all MCMC iterations, provides an estimate of the probability that the two time series belong to the same cluster. Then, in order to compare among cluster structures resulting from different model specifications, we use the following heterogeneity measure

where

denotes the set of indices for a structure of

clusters with size-groups

.…”

Section: Clustering Time Series Of Confirmed Covid-19 Deathsmentioning

confidence: 99%

Clustering spatio-temporal series of confirmed COVID-19 deaths in Europe

et al. 2022

View full text Add to dashboard Cite

“…• Time Series Data: Connected with the topic of longitudinal data is also that of time series data. In this direction some recently proposed models include Nieto- Barajas et al (2012), Jara et al (2013), and Nieto- Barajas et al (2014).…”

Section: Final Remarksmentioning

confidence: 99%

Bayesian Nonparametric Biostatistics

Johnson

Carvalho

2015

Nonparametric Bayesian Inference in Biostatistics

View full text Add to dashboard Cite

We discuss some typical applications of Bayesian nonparametrics in biostatistics. The chosen applications highlight how Bayesian nonparametrics can contribute to addressing some fundamental questions that arise in biomedical research. In particular, we review some modern Bayesian semi-and nonparametric approaches for modeling longitudinal, survival, and medical diagnostic outcome data. Our discussion includes methods for longitudinal data analysis, non-proportional hazards survival analysis, joint modeling of longitudinal and survival data, longitudinal diagnostic test outcome data, and receiver operating characteristic curves. Throughout, we make comparisons among competing BNP models for the various data types considered.

show abstract

A Bayesian Nonparametric Approach for Time Series Clustering

Cited by 37 publications

References 29 publications

Analysing the health effects of simultaneous exposure to physical and chemical properties of airborne particles

Analysing the health effects of simultaneous exposure to physical and chemical properties of airborne particles

Clustering spatio-temporal series of confirmed COVID-19 deaths in Europe

Bayesian Nonparametric Biostatistics

Contact Info

Product

Resources

About