Model-Based Clustering of Trends and Cycles of Nitrate Concentrations in Rivers Across France

Camille Roland Marie Minaudo
SPRINGER, 2022
Article

Résumé

Elevated nitrate from human activity causes ecosystem and economic harm globally. The factors that control the spatiotemporal dynamics of riverine nitrate concentration remain difficult to describe and predict. We analyzed nitrate concentration from 4450 sites throughout France to group sites that exhibit similar trend and seasonal behaviors during 2010-2017 and relate these dynamics to catchment characteristics. We employed a latent-variable, Bayesian mixture of harmonic regressions model to infer site clustering based on multi-year trend and annual cycle amplitude and phase. We examined clustering patterns and relationships among nitrate level, trend, and seasonality parameters. Cluster membership probabilities were governed by continuous, latent variables that were informed with seven classes of covariates encompassing geology, hydrology, and land use. To relate interpretable parameters to the covariates, we modeled amplitude and phase separately in a novel framework employing a bivariate phase regression with the projected normal distribution. The analysis identified regional regimes of nitrate dynamics, including trend classifications. This approach can reveal general patterns that transcend small-scale heterogeneity, complementing site-level assessments to inform regional- to national-level progress in water quality. Supplementary materials accompanying this paper appear on-line.

Official source

https://infoscience.epfl.ch/record/296618?ln=fr

À propos de ce résultat

Cette page est générée automatiquement et peut contenir des informations qui ne sont pas correctes, complètes, à jour ou pertinentes par rapport à votre recherche. Il en va de même pour toutes les autres pages de ce site. Veillez à vérifier les informations auprès des sources officielles de l'EPFL.

Concepts associés

Chargement

Publications associées

Chargement

We introduce a fast approach to classification and clustering applicable to high-dimensional continuous data, based on Bayesian mixture models for which explicit computations are available. This permits us to treat classification and clustering in a single framework, and allows calculation of unobserved class probability. The new classifier is robust to adding noise variables as a drawback of the built-in spike-and-slab structure of the proposed Bayesian model. The usefulness of classification using our method is shown on metabololomic example, and on the Iris data with and without noise variables. Agglomerative hierarchical clustering is used to construct a dendrogram based on the posterior probabilities of particular partitions, to provide a dendrogram with a probabilistic interpretation. An extension to variable selection is proposed which summarises the importance of variables for classification or clustering and has probabilistic interpretation. Having a simple model provides estimation of the model parameters using maximum likelihood and therefore yields a fully automatic algorithm. The new clustering method is applied to metabolomic, microarray, and image data and is studied using simulated data motivated by real datasets. The computational difficulties of the new approach are discussed, solutions for algorithm acceleration are proposed, and the written computer code is briefly analysed. Simulations shows that the quality of the estimated model parameters depends on the parametric distribution assumed for effects, but after fixing the model parameters to reasonable values, the distribution of the effects influences clustering very little. Simulations confirms that the clustering algorithm and the proposed variable selection method is reliable when the model assumptions are wrong. The new approach is compared with the popular Bayesian clustering alternative, MCLUST, fitted on the principal components using two loss functions in which our proposed approach is found to be more efficient in almost every situation.

EPFL2009

Chargement

Personalizable intervention systems to promote healthy behavior change

Igor Kulev

Adopting healthy behaviors can prevent the onset of many adverse health conditions. However, behavior changes are difficult to make, and often, people who like to improve their behaviors do not know how to do that. Personalizable intervention systems could assist them to achieve healthy behavior change. These systems decide what would be the optimal intervention for the target user based on his or her characteristics, including current and past behavior patterns. In this thesis, we propose novel solutions that address the main challenges in building a personalizable intervention system to promote healthy behavior change. First, we propose a system based on a Bayesian mixture model to identify subpopulations with different behavior changes from longitudinal data. This system is especially suitable when the amount of data is limited, and when there are unobserved factors that might affect behavior change. Second, we propose CLINT, a system based on a latent-variable model, to discover and predict behavior change patterns from fine-grained sensor data. The novelty of this system is that it produces interpretable patterns that could be used to suggest successful behavior change strategies from the existing users similar to the target user. Third, we propose a personalizable intervention system to improve the physical activeness of senior adults. The main novelty of this system is that it uses historical time series fitness data to decide which intervention to recommend. Finally, we propose ACFR, an adversarial approach to reduce intervention bias in observational data. This approach learns a balanced representation of the covariates that allows personalizable intervention systems to make a better estimate of the intervention effect. Our solutions turn existing human behavior data into actionable insights for future users who may have unhealthy lifestyles.

EPFL2020

Chargement

Un modèle factoriel dynamique pour séries temporelles

Andrei Zenide

This work deals with factorial models for multiple time series. Its core content puts it at the interface between statistics and finance. After a brief description of the historical link between the two sciences, it reviews the literature on factorial models that are close to the model introduced in this work called The dynamical factor analysis model for time series. This model makes the hypothesis that the observed time series are influenced by a common factor, difficult to define and impossible to measure. No a priori structure is put on the factor, at each point time the value of the factor is considered as a new parameter that has to be estimated. As a consequence of this fact, the number of parameters is large and it is not possible to provide the usual asimptotic properties of the estimations by letting the number of periods tend to infi- nity. Asymptotic results in our context have been obtained by increasing the number of time series. The model makes the hypothesis that there is a linear dependence between the time series and the factor with coefficients, that are not constant over time, but rather follow a smooth random walk. This mean that the linear structure of the model is evolving slowly from a period to another. All the information which is not contained in the factor and the coefficients is considered as white noise. Using the normal distribution makes the estimation more easy and opens a toolbox of statistical methods that have been developed for this kind of data. The model is a part of the family of state space models and the Kalman filter is an essential ingredient of our estimator. The effort was concentrated on the elaboration of the structure of the model. Its complexity was constrained by the difficulty of estimation. The final shape of the model does not allow an analytical solution of the optimization problem introduced by the maximum likelihood estimation of the parameters. Numerical solutions have been found and compared with the parameters for simulated data. Some others models have been developed as simpler versions of the dynamical factor model for time series. The case where the factor can be observed has been studied and a new method for the estimation has been provided and compared with the existing methods. A second study considers a latent factor model without noise. Two methods for the estimation of the factor have been provided. The last chapter contains a detailed description of the main statistical tools used during this work. The links with the previous chapters are highlighted followed by comments.

EPFL2004

Chargement

EPFL2009

Un modèle factoriel dynamique pour séries temporelles

Andrei Zenide

EPFL2004