Calibrating covariate informed product partition models

Covariate informed product partition models incorporate the intuitively appealing notion that individuals or units with similar covariate values a priori have a higher probability of co-clustering than those with dissimilar covariate values. These methods have been shown to perform well if the number of covariates is relatively small. However, as the number of covariates increase, their influence on partition probabilities overwhelm any information the response may provide in clustering and often encourage partitions with either a large number of singleton clusters or one large cluster resulting in poor model fit and poor out-of-sample prediction. This same phenomenon is observed in Bayesian nonparametric regression methods that induce a conditional distribution for the response given covariates through a joint model. In light of this, we propose two methods that calibrate the covariate-dependent partition model by capping the influence that covariates have on partition probabilities. We demonstrate the new methods’ utility using simulation and two publicly available datasets.     

Bayesian clustering and product partition models

Summary. We present a decision theoretic formulation of product partition models (PPMs) that allows a formal treatment of different decision problems such as estimation or hypothesis testing and clustering methods simultaneously. A key observation in our construction is the fact that PPMs can be formulated in the context of model selection. The underlying partition structure in these models is closely related to that arising in connection with Dirichlet processes. This allows a straightforward adaptation of some computational strategies—originally devised for nonparametric Bayesian problems—to our framework. The resulting algorithms are more flexible than other competing alternatives that are used for problems involving PPMs. We propose an algorithm that yields Bayes estimates of the quantities of interest and the groups of experimental units. We explore the application of our methods to the detection of outliers in normal and Student t regression models, with clustering structure equivalent to that induced by a Dirichlet process prior. We also discuss the sensitivity of the results considering different prior distributions for the partitions.     

Random partition models with regression on covariates

Many recent applications of nonparametric Bayesian inference use random partition models, i.e. probability models for clustering a set of experimental units. We review the popular basic constructions. We then focus on an interesting extension of such models. In many applications covariates are available that could be used to a priori inform the clustering. This leads to random clustering models indexed by covariates, i.e., regression models with the outcome being a partition of the experimental units. We discuss some alternative approaches that have been used in the recent literature to implement such models, with an emphasis on a recently proposed extension of product partition models. Several of the reviewed approaches were not originally intended as covariate-based random partition models, but can be used for such inference.     

Bayesian modeling using a class of bimodal skew-elliptical distributions

We consider Bayesian inference using an extension of the family of skew-elliptical distributions studied by Azzalini [1985. A class of distributions which includes the normal ones. Scand. J. Statist. Theory and Applications 12 (2), 171–178]. This new class is referred to as bimodal skew-elliptical (BSE) distributions. The elements of the BSE class can take quite different forms. In particular, they can adopt both uni- and bimodal shapes. The bimodal case behaves similarly to mixtures of two symmetric distributions and we compare inference under the BSE family with the specific case of mixtures of two normal distributions. We study the main properties of the general class and illustrate its applications to two problems involving density estimation and linear regression.     

Economic recession and the reverse of internal migration flows of Latin American immigrants in Spain

Fluctuations in internal migration flows have been closely associated with changes in economic processes. However, foreign immigrants’ spatial responses to changing national economic conditions have not yet been sufficiently studied. This paper aims to analyse the impact of the Spanish economic downturn on Latin American immigrants’ mobility by examining their territorial patterns of internal migration. In order to elucidate the intensity and directionality of Latin Americans’ internal flows between 2004 and 2013, divided in two periods – before and during the crisis–, we relied on the data of the Residential Variation Statistics, and the creation of a typology of provinces according to each province’s socio-economic characteristics. We calculated the net migration rates for the main immigrant groups and generated origin–destination matrices. This study shows that Latin Americans’ internal flows in Spain have been affected by the economic downturn, revealing a decrease in intensity and a change in directionality. The economic crisis in Spain has caused a change in the previous dynamic of geographical dispersion and has become a new factor of differentiation in the distribution of the Latin American population in Spain.     

False alarms and incorrect rejections in an information security center: correlation with the frequency of incidents

This study examines the actions taken by operators aimed at preventing and combating information security incidents at a banking organization. The work utilizes the theoretical framework of ergonomics and cognitive psychology. The method is workplace ergonomic analysis. Its focus is directed towards examining the cognitive dimension of the work environment with special attention to the occurrence of correlations between variability in incident frequency and the results of sign detection actions. It categorizes 45,142 operator decisions according to the theory of signal detection (Sternberg, 2000). It analyzes the correlation between incident proportions (indirectly associated with the cognitive efforts demanded from the operator) and operator decisions. The study demonstrated the existence of a positive correlation between incident proportions and false positive decisions (false alarms). However, this correlation could not be observed in relation to decisions of the false-negative type (incorrect rejection).     

A field experiment comparing mechanical demands of two pruners for flower cutting

Manual flower cutting requires repetitive forceful movements that may explain the high incidence of upper-extremity MSDs among workers of the Colombian flower industry. Force demands may be caused by the use of pruners that have not been designed for Colombian workers. We conducted a repeated measures field experimental study (n=16, 69% women, 20 to 60 years old) to compare the mechanical demands during rose cutting of two different pruners (Traditional vs. alternative pruner) at two different cutting heights (elbow vs. over-shoulder). Demands included the perceived comfort, perceived effort and upper limb's muscular activity and posture. Workers perceived their traditional pruner as more comfortable but requiring more effort than the prototype tool. No benefits were observed regarding ulnar deviation with the alternative pruner. The wrist extension and related muscular activity of the wrist extensors were higher with the prototype pruner. The prototype pruner resulted in reduced muscular load of the wrist flexors; the biceps when working at the elbow height; and the deltoids when working at the shoulder height. However, Moderate results may be explained by the lack of time for workers' adaptation to the alternative tool. New design developments must focus on reducing extension and ulnar deviation of the wrist.