Mixture of bivariate Poisson regression models with an application to insurance

In a recent paper Bermúdez [2009] used bivariate Poisson regression models for ratemaking in car insurance, and included zero-inflated models to account for the excess of zeros and the overdispersion in the data set. In the present paper, we revisit this model in order to consider alternatives. We propose a 2-finite mixture of bivariate Poisson regression models to demonstrate that the overdispersion in the data requires more structure if it is to be taken into account, and that a simple zero-inflated bivariate Poisson model does not suffice. At the same time, we show that a finite mixture of bivariate Poisson regression models embraces zero-inflated bivariate Poisson regression models as a special case. Additionally, we describe a model in which the mixing proportions are dependent on covariates when modelling the way in which each individual belongs to a separate cluster. Finally, an EM algorithm is provided in order to ensure the models’ ease-of-fit. These models are applied to the same automobile insurance claims data set as used in Bermúdez [2009] and it is shown that the modelling of the data set can be improved considerably.

Comment on "Neutron brillouin scattering study of collective dynamics in a dense He-Ne gaseous mixture"

A Comment on the Letter by Ubaldo Bafile, et al., Phys. Rev. Lett. 86, 1019 (2001). The authors of the Letter offer a Reply.

Contribution of microbial and invertebrate communities to leaf litter colonization in a Mediterranean stream

Leaf litter inputs and retention play an important role in ecosystem functioning in forested streams. We examined colonization of leaves by microbes (bacteria, fungi, and protozoa) and fauna in Fuirosos, an intermittent forested Mediterranean stream. Black poplar (Populus nigra) and plane (Platanus acerifolia) leaf packs were placed in the stream for 4 mo. We measured the biomasses and calculated the densities of bacteria, fungi, protozoa, meiofauna, and macroinvertebrates to determine their dynamics and potential interactions throughout the colonization process. Colonization was strongly correlated with hydrological variability (defined mainly by water temperature and discharge). The 1st week of colonization was characterized by hydrological stability and warm water temperatures, and allocation of C from microbial to invertebrate compartments on the leaf packs was rapid. Clumps of fine particulate organic matter (FPOM) were retained by the leaf packs, and enhanced rapid colonization by microfauna and meiofaunal collector-gatherers (ostracods and copepods). After 2 wk, an autumnal flood caused a 20-fold increase in water flow. Higher discharge and lower water temperature caused FPOM-related fauna to drift away from the packs and modified the subsequent colonization sequence. Fungi showed the highest biomass, with similar values to those recorded at the beginning of the experiment. After 70 d of postflood colonization, fungi decreased to nearly 40% of the total C in the leaf packs, whereas invertebrates became more abundant and accounted for 60% of the C. Natural flood occurrence in Mediterranean streams could be a key factor in the colonization and processing of organic matter.

The effect of pesticides and alien invesive species on soil biota and litter decomposition rates in a Mediterrean-climate ecosystem of Western Australia

Ecosystems are complex systems and changing one of their components can alter their whole functioning. Decomposition and biodiversity are two factors that play a role in this stability, and it is vital to study how these two factors are interrelated and how other factors, whether of human origin or not, can affect them. This study has tested different hypotheses regarding the effects of pesticides and invasive species on the biodiversity of the soil fauna and litter decomposition rate. Decomposition was measured using the litterbags technique. Our results indicate that pesticides had a negative effect on decomposition whereas invasive species increased decomposition rate. At the same time, the diversity of the soil biota was unaffected by either factor. These results allow us to better understand the response of important ecosystem functions to human‐induced alterations, in order to mitigate harmful effects or restore them wherever necessary.

Finite mixture analysis of beauty-contest data using generalised beta distributions

This paper introduces a mixture model based on the beta distribution, without preestablishedmeans and variances, to analyze a large set of Beauty-Contest data obtainedfrom diverse groups of experiments (Bosch-Domenech et al. 2002). This model gives a bettert of the experimental data, and more precision to the hypothesis that a large proportionof individuals follow a common pattern of reasoning, described as iterated best reply (degenerate),than mixture models based on the normal distribution. The analysis shows thatthe means of the distributions across the groups of experiments are pretty stable, while theproportions of choices at dierent levels of reasoning vary across groups.

Technical Note: Dissolved organic matter fluorescence a finite mixture approach to deconvolve excitation-emission matrices

The analysis of the shape of excitation-emission matrices (EEMs) is a relevant tool for exploring the origin, transport and fate of dissolved organic matter (DOM) in aquatic ecosystems. Within this context, the decomposition of EEMs is acquiring a notable relevance. A simple mathematical algorithm that automatically deconvolves individual EEMs is described, creating new possibilities for the comparison of DOM fluorescence properties and EEMs that are very different from each other. A mixture model approach is adopted to decompose complex surfaces into sub-peaks. The laplacian operator and the Nelder-Mead optimisation algorithm are implemented to individuate and automatically locate potential peaks in the EEM landscape. The EEMs of a simple artificial mixture of fluorophores and DOM samples collected in a Mediterranean river are used to describe the model application and to illustrate a strategy that optimises the search for the optimal output.

Improving algorithm speed in PNL mixture separation and Wiener system inversion

This paper proposes a very simple method for increasing the algorithm speed for separating sources from PNL mixtures or invertingWiener systems. The method is based on a pertinent initialization of the inverse system, whose computational cost is very low. The nonlinear part is roughly approximated by pushing the observations to be Gaussian; this method provides a surprisingly good approximation even when the basic assumption is not fully satisfied. The linear part is initialized so that outputs are decorrelated. Experiments shows the impressive speed improvement.

Interactions of bacteria and fungi on decomposing litter: differential extracellular enzyme activities

Fungi and bacteria are key agents in plant litter decomposition in freshwater ecosystems. However, the specific roles of these two groups and their interactions during the decomposition process are unclear. We compared the growth and patterns of degradativeenzymes expressed by communities of bacteria and fungi grown separately and in coexistence on Phragmites leaves. The two groups displayed both synergistic and antagonistic interactions. Bacteria grew better together with fungi than alone. In addition, there was a negative effect of bacteria on fungi, which appeared to be caused by suppression of fungal growth and biomass accrual rather than specifically affecting enzyme activity. Fungi growing alone had a high capacity for the decomposition of plant polymers such as lignin, cellulose, and hemicellulose. In contrast, enzyme activities were in general low when bacteria grew alone, and the activity of key enzymes in the degradation of lignin and cellulose (phenol oxidase and cellobiohydrolase) was undetectable in the bacteria-only treatment. Still, biomass-specific activities of most enzymes were higher in bacteria than in fungi. The low total activity and growth of bacteria in the absence of fungi in spite of apparent high enzymatic efficiency during the degradation of many substrates suggest that fungi provide the bacteria with resources that the bacteria were not able to acquire on their own, most probably intermediate decomposition products released by fungi that could be used by bacteria

Sabbatical stay at CREVER, University Rovira i Virgili

A three-year research proposal of international impact and quality was developed between the author and CREVER researchers, taking into account the infrastructure and research lines already being carried out at CREVER. The project is related to the study of new mixtures in thermal activated systems, by the addition of a third component to help the heat transfer processes. The project proposes the use of water and lithium nitrate as absorbents in ternary ammonia mixtures, varying the concentration with the objective of optimising the mixture for solar air conditioning purposes. Also, the research proposal will promote an intensive collaboration in the following years between CREVER and The Centro de Investigación en Energía-UNAM, Mexico.

Modeling usage of medical care services: the medical expenditure panel survey data, 1996-2000

We explore the determinants of usage of six different types of health care services, using the Medical Expenditure Panel Survey data, years 1996-2000. We apply a number of models for univariate count data, including semiparametric, semi-nonparametric and finite mixture models. We find that the complexity of the model that is required to fit the data well depends upon the way in which the data is pooled across sexes and over time, and upon the characteristics of the usage measure. Pooling across time and sexes is almost always favored, but when more heterogeneous data is pooled it is often the case that a more complex statistical model is required.

Polarization, fractionalization and conflict

A recent upsurge of empirical studies on the causes of conflict attempts to connect various features of the distribution of the relevant characteristic (typically ethnicity or religion) to conflict. The distributional indices differ (polarization, fractionalization or Lorenz-domination) and so do the various specifications of "conflict" (onset, incidence or intensity). Overall, the results are far from clear, and combined with the mixture of alternative indices and notions of "conflict" it is not surprising that the reader may come away thoroughly perplexed. The aim of this paper is to provide a theoretical framework that permits us to distinguish between the occurrence of conflict and its severity and that clarifies the role of polarization and fractionalization in each of these cases.

Occurrence and bioaccumulation study of PCDD and PCDF from mineral feed additives

Occurrence of polychlorinated dibenzo-p-dioxins (PCDDs) and polychlorinated dibenzofurans (PCDFs) was evaluated in sepiolite as a widely employed binder and anti-caking agent for animal feed. Also, naturally contaminated kaolinitic clay was used for comparative purposes. Since sepiolite shows remarkable adsorption properties, particular interest was paid to the extraction steps as they become critical for the final determination of these pollutants in such matrixes. Furthermore, classical Soxhlet extraction using different extracting strategies as well as acid treatment were carried out with simultaneous liquid-liquid extraction. Results obtained depended on the extraction procedure applied. Acid treatment or Soxhlet extraction using a mixture of toluene:ethanol as solvent allowed to reach the minimum requirements of recovery rates. However, Soxhlet extraction using a mixture cyclohexane:toluene as extracting solvent did not allow to comply with minimum specifications for recovery. Significant differences were obtained in TEQ units when acid treatment was applied in comparison to Soxhlet extraction. This fact can be explained because the use of drastic acid conditions allows removing strongly adsorbed analytes which can be uniquely extracted after a total destruction of the crystalline structure of sepiolite. On the contrary, Soxhlet extraction was not able to destroy the structure of sepiolite and as a consequence the PCDDs/Fs were strongly adsorbed in the internal structure of the mineral. From biological point of view the availability of these toxicants constitutes a critical aspect playing an important role in the final decision choosing particular analytical procedures. Then, acid conditions in the digestive tract should be taken into account. In this scenario, a bioaccumulation study was conducted to evaluate the transference of PCDDs/PCDFs from the sepiolite into the animal tissues when fed with feed containing sepiolite. To this end, chickens were used as a model to examine the bioavailability of PCDDs/PCDFs. Four groups of chickens were exposed through their diet to a control feed, feed with 3% w/w sepiolite as additive, feed contaminated with PCDDs/PCDFs at concentration around 2.8 pg WHO-TEQ/g and feed with 2% of a contaminated kaolinitic clay (460 pg TEQ/g mineral). Livers of the four studied groups were analyzed throughout the exposure period. Results of this trial showed that the performance of broilers was not affected by the presence of dioxins at levels tested, and chickens did not show any abnormal behaviour. Dioxins intentionally added to the diet were absorbed and accumulated in the liver in a significant manner, whereas the PCDDs/Fs from sepiolite were not available for chickens since livers from broilers fed 3% sepiolite presented similar WHO-TEQ values than those from broilers fed control diet.