Using decision trees to understand structure in missing data

Objectives Demonstrate the application of decision trees – classification and regression trees (CARTs), and their cousins, boosted regression trees (BRTs) – to understand structure in missing data. Setting Data taken from employees at three different industry sites in Australia. Participants 7915 observations were included. Materials and Methods The approach was evaluated using an occupational health dataset comprising results of questionnaires, medical tests, and environmental monitoring. Statistical methods included standard statistical tests and the ‘rpart’ and ‘gbm’ packages for CART and BRT analyses, respectively, from the statistical software ‘R’. A simulation study was conducted to explore the capability of decision tree models in describing data with missingness artificially introduced. Results CART and BRT models were effective in highlighting a missingness structure in the data, related to the Type of data (medical or environmental), the site in which it was collected, the number of visits and the presence of extreme values. The simulation study revealed that CART models were able to identify variables and values responsible for inducing missingness. There was greater variation in variable importance for unstructured compared to structured missingness. Discussion Both CART and BRT models were effective in describing structural missingness in data. CART models may be preferred over BRT models for exploratory analysis of missing data, and selecting variables important for predicting missingness. BRT models can show how values of other variables influence missingness, which may prove useful for researchers. Conclusion Researchers are encouraged to use CART and BRT models to explore and understand missing data.

Threshold detection: matching statistical methodology to ecological questions and conservation planning objectives

Two types of ecological thresholds are now being widely used to develop conservation targets: breakpoint-based thresholds represent tipping points where system properties change dramatically, whereas classification thresholds identify groups of data points with contrasting properties. Both breakpoint-based and classification thresholds are useful tools in evidence-based conservation. However, it is critical that the type of threshold to be estimated corresponds with the question of interest and that appropriate statistical procedures are used to determine its location. On the basis of their statistical properties, we recommend using piecewise regression methods to identify breakpoint-based thresholds and discriminant analysis or classification and regression trees to identify classification thresholds.

Effects of basic fibroblast growth factor on density and morphology of fibroblasts grown on root surfaces with or without conditioning with tetracycline or EDTA.

A study was conducted to evaluate in vitro the effect of root surface conditioning with basic fibroblast growth factor (b-FGF) on morphology and proliferation of fibroblasts. Three experimental groups were used: non-treated, and treated with 50 microg or 125 microg b-FGF/ml. The dentin samples in each group were divided into subgroups according to the chemical treatment received before application of b-FGF: none, or conditioned with tetracycline-HCl or EDTA. After contact with b-FGF for 5 min, the samples were incubated for 24 h with 1 ml of culture medium containing 1 x 10(5) cells/ml plus 1 ml of culture medium alone. The samples were then subjected to routine preparation for SEM, and random fields were photographed. Three calibrated and blind examiners performed the assessment of morphology and density according to two index systems. Classification and regression trees indicated that the root surfaces treated with 125 microg b-FGF and previously conditioned with tetracycline-HCl or EDTA presented a morphology more suggestive of cellular adhesion and viability (P = 0.004). The density of fibroblasts on samples previously conditioned with EDTA, regardless of treatment with b-FGF, was significantly higher than in the other groups (P < 0.001). The present findings suggest that topical application of b-FGF has a positive influence on both the density and morphology of fibroblasts.

Life-span, T-cell responses, and incidence of lymphomas in congenic mice.

Survival, T-cell functions, and postmortem histopathology were studied in H-2 congenic strains of mice bearing H-2b, H-2k, and H-2d haplotypes. Males lived longer than females in all homozygous and heterozygous combinations except for H-2d homozygotes, which showed no differences between males and females. Association of heterozygosity with longer survival was observed only with H-2b/H-2b and H-2b/H-2d mice. Analysis using classification and regression trees (CART) showed that both males and females of H-2b homozygous and H-2k/H-2b mice had the shortest life-span of the strains studied. In histopathological analyses, lymphomas were noted to be more frequent in females, while hemangiosarcomas and hepatomas were more frequent in males. Lymphomas appeared earlier than hepatomas or hemangiosarcomas. The incidence of lymphomas was associated with the H-2 haplotype--e.g., H-2b homozygous mice had more lymphomas than did mice of the H-2d haplotype. More vigorous T-cell function was maintained with age (27 months) in H-2d, H-2b/H-2d, and H-2d/H-2k mice as compared with H-2b, H-2k, and H-2b/H-2k mice, which showed a decline of T-cell responses with age.

Modelling pre-clearing vegetation distribution using GIS-integrated statistical, ecological and data models: A case study from the wet tropics of Northeastern Australia

Traditional vegetation mapping methods use high cost, labour-intensive aerial photography interpretation. This approach can be subjective and is limited by factors such as the extent of remnant vegetation, and the differing scale and quality of aerial photography over time. An alternative approach is proposed which integrates a data model, a statistical model and an ecological model using sophisticated Geographic Information Systems (GIS) techniques and rule-based systems to support fine-scale vegetation community modelling. This approach is based on a more realistic representation of vegetation patterns with transitional gradients from one vegetation community to another. Arbitrary, though often unrealistic, sharp boundaries can be imposed on the model by the application of statistical methods. This GIS-integrated multivariate approach is applied to the problem of vegetation mapping in the complex vegetation communities of the Innisfail Lowlands in the Wet Tropics bioregion of Northeastern Australia. The paper presents the full cycle of this vegetation modelling approach including sampling sites, variable selection, model selection, model implementation, internal model assessment, model prediction assessments, models integration of discrete vegetation community models to generate a composite pre-clearing vegetation map, independent data set model validation and model prediction's scale assessments. An accurate pre-clearing vegetation map of the Innisfail Lowlands was generated (0.83r(2)) through GIS integration of 28 separate statistical models. This modelling approach has good potential for wider application, including provision of. vital information for conservation planning and management; a scientific basis for rehabilitation of disturbed and cleared areas; a viable method for the production of adequate vegetation maps for conservation and forestry planning of poorly-studied areas. (c) 2006 Elsevier B.V. All rights reserved.

Higher-order co-occurrences for exploratory point pattern analysis and decision tree clustering on spatial data

Analyzing geographical patterns by collocating events, objects or their attributes has a long history in surveillance and monitoring, and is particularly applied in environmental contexts, such as ecology or epidemiology. The identification of patterns or structures at some scales can be addressed using spatial statistics, particularly marked point processes methodologies. Classification and regression trees are also related to this goal of finding "patterns" by deducing the hierarchy of influence of variables on a dependent outcome. Such variable selection methods have been applied to spatial data, but, often without explicitly acknowledging the spatial dependence. Many methods routinely used in exploratory point pattern analysis are2nd-order statistics, used in a univariate context, though there is also a wide literature on modelling methods for multivariate point pattern processes. This paper proposes an exploratory approach for multivariate spatial data using higher-order statistics built from co-occurrences of events or marks given by the point processes. A spatial entropy measure, derived from these multinomial distributions of co-occurrences at a given order, constitutes the basis of the proposed exploratory methods. © 2010 Elsevier Ltd.

Higher-order co-occurrences for exploratory point pattern analysis and decision tree clustering on spatial data

Analyzing geographical patterns by collocating events, objects or their attributes has a long history in surveillance and monitoring, and is particularly applied in environmental contexts, such as ecology or epidemiology. The identification of patterns or structures at some scales can be addressed using spatial statistics, particularly marked point processes methodologies. Classification and regression trees are also related to this goal of finding "patterns" by deducing the hierarchy of influence of variables on a dependent outcome. Such variable selection methods have been applied to spatial data, but, often without explicitly acknowledging the spatial dependence. Many methods routinely used in exploratory point pattern analysis are2nd-order statistics, used in a univariate context, though there is also a wide literature on modelling methods for multivariate point pattern processes. This paper proposes an exploratory approach for multivariate spatial data using higher-order statistics built from co-occurrences of events or marks given by the point processes. A spatial entropy measure, derived from these multinomial distributions of co-occurrences at a given order, constitutes the basis of the proposed exploratory methods. © 2010 Elsevier Ltd.

Influence of environmental variability and age on the body condition of small pelagic fish in the Gulf of Lions

Endogenous and environmental variables are fundamental in explaining variations in fish condition. Based on more than 20 yr of fish weight and length data, relative condition indices were computed for anchovy and sardine caught in the Gulf of Lions. Classification and regression trees (CART) were used to identify endogenous factors affecting fish condition, and to group years of similar condition. Both species showed a similar annual cycle with condition being minimal in February and maximal in July. CART identified 3 groups of years where the fish populations generally showed poor, average and good condition and within which condition differed between age classes but not according to sex. In particular, during the period of poor condition (mostly recent years), sardines older than 1 yr appeared to be more strongly affected than younger individuals. Time-series were analyzed using generalized linear models (GLMs) to examine the effects of oceanographic abiotic (temperature, Western Mediterranean Oscillation [WeMO] and Rhone outflow) and biotic (chlorophyll a and 6 plankton classes) factors on fish condition. The selected models explained 48 and 35% of the variance of anchovy and sardine condition, respectively. Sardine condition was negatively related to temperature but positively related to the WeMO and mesozooplankton and diatom concentrations. A positive effect of mesozooplankton and Rhone runoff on anchovy condition was detected. The importance of increasing temperatures and reduced water mixing in the NW Mediterranean Sea, affecting planktonic productivity and thus fish condition by bottom-up control processes, was highlighted by these results. Changes in plankton quality, quantity and phenology could lead to insufficient or inadequate food supply for both species.

Accurately defining the reference condition for summary biotic metrics: A comparison of four approaches

Protocols for bioassessment often relate changes in summary metrics that describe aspects of biotic assemblage structure and function to environmental stress. Biotic assessment using multimetric indices now forms the basis for setting regulatory standards for stream quality and a range of other goals related to water resource management in the USA and elsewhere. Biotic metrics are typically interpreted with reference to the expected natural state to evaluate whether a site is degraded. It is critical that natural variation in biotic metrics along environmental gradients is adequately accounted for, in order to quantify human disturbance-induced change. A common approach used in the IBI is to examine scatter plots of variation in a given metric along a single stream size surrogate and a fit a line (drawn by eye) to form the upper bound, and hence define the maximum likely value of a given metric in a site of a given environmental characteristic (termed the 'maximum species richness line' - MSRL). In this paper we examine whether the use of a single environmental descriptor and the MSRL is appropriate for defining the reference condition for a biotic metric (fish species richness) and for detecting human disturbance gradients in rivers of south-eastern Queensland, Australia. We compare the accuracy and precision of the MSRL approach based on single environmental predictors, with three regression-based prediction methods (Simple Linear Regression, Generalised Linear Modelling and Regression Tree modelling) that use (either singly or in combination) a set of landscape and local scale environmental variables as predictors of species richness. We compared the frequency of classification errors from each method against set biocriteria and contrast the ability of each method to accurately reflect human disturbance gradients at a large set of test sites. The results of this study suggest that the MSRL based upon variation in a single environmental descriptor could not accurately predict species richness at minimally disturbed sites when compared with SLR's based on equivalent environmental variables. Regression-based modelling incorporating multiple environmental variables as predictors more accurately explained natural variation in species richness than did simple models using single environmental predictors. Prediction error arising from the MSRL was substantially higher than for the regression methods and led to an increased frequency of Type I errors (incorrectly classing a site as disturbed). We suggest that problems with the MSRL arise from the inherent scoring procedure used and that it is limited to predicting variation in the dependent variable along a single environmental gradient.

Methods in general model localization

The aim of this study was to evaluate and test methods which could improve local estimates of a general model fitted to a large area. In the first three studies, the intention was to divide the study area into sub-areas that were as homogeneous as possible according to the residuals of the general model, and in the fourth study, the localization was based on the local neighbourhood. According to spatial autocorrelation (SA), points closer together in space are more likely to be similar than those that are farther apart. Local indicators of SA (LISAs) test the similarity of data clusters. A LISA was calculated for every observation in the dataset, and together with the spatial position and residual of the global model, the data were segmented using two different methods: classification and regression trees (CART) and the multiresolution segmentation algorithm (MS) of the eCognition software. The general model was then re-fitted (localized) to the formed sub-areas. In kriging, the SA is modelled with a variogram, and the spatial correlation is a function of the distance (and direction) between the observation and the point of calculation. A general trend is corrected with the residual information of the neighbourhood, whose size is controlled by the number of the nearest neighbours. Nearness is measured as Euclidian distance. With all methods, the root mean square errors (RMSEs) were lower, but with the methods that segmented the study area, the deviance in single localized RMSEs was wide. Therefore, an element capable of controlling the division or localization should be included in the segmentation-localization process. Kriging, on the other hand, provided stable estimates when the number of neighbours was sufficient (over 30), thus offering the best potential for further studies. Even CART could be combined with kriging or non-parametric methods, such as most similar neighbours (MSN).

Lahopuumäärän ennustaminen ja kartoitus lentokonelaserkeilauksen avulla

Lahopuun määrästä ja sijoittumisesta ollaan kiinnostuneita paitsi elinympäristöjen monimuotoisuuden, myös ilmakehän hiilen varastoinnin kannalta. Tutkimuksen tavoitteena oli kehittää aluepohjainen laserkeilausdataa hyödyntävä malli lahopuukohteiden paikantamiseksi ja lahopuun määrän estimoimiseksi. Samalla tutkittiin mallin selityskyvyn muuttumista mallinnettavan ruudun kokoa suurennettaessa. Tutkimusalue sijaitsi Itä-Suomessa Sonkajärvellä ja koostui pääasiassa nuorista hoidetuista talousmetsistä. Tutkimuksessa käytettiin harvapulssista laserkeilausdataa sekä kaistoittain mitattua maastodataa kuolleesta puuaineksesta. Aineisto jaettiin siten, että neljäsosa datasta oli käytössä mallinnusta varten ja loput varattiin valmiiden mallien testaamiseen. Lahopuun mallintamisessa käytettiin sekä parametrista että ei-parametrista mallinnusmenetelmää. Logistisen regression avulla erikokoisille (0,04, 0,20, 0,32, 0,52 ja 1,00 ha) ruuduille ennustettiin todennäköisyys lahopuun esiintymiselle. Muodostettujen mallien selittävät muuttujat valittiin 80 laserpiirteen ja näiden muunnoksien joukosta. Mallien selittävät muuttujat valittiin kolmessa vaiheessa. Aluksi muuttujia tarkasteltiin visuaalisesti kuvaamalla ne lahopuumäärän suhteen. Ensimmäisessä vaiheessa sopivimmiksi arvioitujen muuttujien selityskykyä testattiin mallinnuksen toisessa vaiheessa yhden muuttujan mallien avulla. Lopullisessa usean muuttujan mallissa selittävien muuttujien kriteerinä oli tilastollinen merkitsevyys 5 % riskitasolla. 0,20 hehtaarin ruutukoolle luotu malli parametrisoitiin muun kokoisille ruuduille. Logistisella regressiolla toteutetun parametrisen mallintamisen lisäksi, 0,04 ja 1,0 hehtaarin ruutukokojen aineistot luokiteltiin ei-parametrisen CART-mallinnuksen (Classification and Regression Trees) avulla. CARTmenetelmällä etsittiin aineistosta vaikeasti havaittavia epälineaarisia riippuvuuksia laserpiirteiden ja lahopuumäärän välillä. CART-luokittelu tehtiin sekä lahopuustoisuuden että lahopuutilavuuden suhteen. CART-luokituksella päästiin logistista regressiota parempiin tuloksiin ruutujen luokituksessa lahopuustoisuuden suhteen. Logistisella mallilla tehty luokitus parani ruutukoon suurentuessa 0,04 ha:sta(kappa 0,19) 0,32 ha:iin asti (kappa 0,38). 0,52 ha:n ruutukoolla luokituksen kappa-arvo kääntyi laskuun (kappa 0,32) ja laski edelleen hehtaarin ruutukokoon saakka (kappa 0,26). CART-luokitus parani ruutukoon kasvaessa. Luokitustulokset olivat logistista mallinnusta parempia sekä 0,04 ha:n (kappa 0,24) että 1,0 ha:n (kappa 0,52) ruutukoolla. CART-malleilla määritettyjen ruutukohtaisten lahopuutilavuuksien suhteellinen RMSE pieneni ruutukoon kasvaessa. 0,04 hehtaarin ruutukoolla koko aineiston lahopuumäärän suhteellinen RMSE oli 197,1 %, kun hehtaarin ruutukoolla vastaava luku oli 120,3 %. Tämän tutkimuksen tulosten perusteella voidaan todeta, että maastossa mitatun lahopuumäärän ja tutkimuksessa käytettyjen laserpiirteiden yhteys on pienellä ruutukoolla hyvin heikko, mutta vahvistuu hieman ruutukoon kasvaessa. Kun mallinnuksessa käytetty ruutukoko kasvaa, pienialaisten lahopuukeskittymien havaitseminen kuitenkin vaikeutuu. Tutkimuksessa kohteen lahopuustoisuus pystyttiin kartoittamaan kohtuullisesti suurella ruutukoolla, mutta pienialaisten kohteiden kartoittaminen ei onnistunut käytetyillä menetelmillä. Pienialaisten kohteiden paikantaminen laserkeilauksen avulla edellyttää jatkotutkimusta erityisesti tiheäpulssisen laserdatan käytöstä lahopuuinventoinneissa.

L’explication de la délinquance prolifique : l’importance de l’interaction entre le risque individuel et le risque contextuel

Le but de cette thèse est d’expliquer la délinquance prolifique de certains délinquants. Nous avançons la thèse que la délinquance prolifique s’explique par la formation plus fréquente de situations criminogènes. Ces situations réfèrent au moment où un délinquant entre en interaction avec une opportunité criminelle dans un contexte favorable au crime. Plus exactement, il s’agit du moment où le délinquant fait face à cette opportunité, mais où le crime n’a pas encore été commis. La formation de situations criminogènes est facilitée par l’interaction et l’interdépendance de trois éléments : la propension à la délinquance de la personne, son entourage criminalisé et son style de vie. Ainsi, la délinquance prolifique ne pourrait être expliquée adéquatement sans tenir compte de l’interaction entre le risque individuel et le risque contextuel. L’objectif général de la présente thèse est de faire la démonstration de l’importance d’une modélisation interactionnelle entre le risque individuel et le risque contextuel afin d’expliquer la délinquance plus prolifique de certains contrevenants. Pour ce faire, 155 contrevenants placés sous la responsabilité de deux établissements des Services correctionnels du Québec et de quatre centres jeunesse du Québec ont complété un protocole d’évaluation par questionnaires auto-administrés. Dans un premier temps (chapitre trois), nous avons décrit et comparé la nature de la délinquance autorévélée des contrevenants de notre échantillon. Ce premier chapitre de résultats a permis de mettre en valeur le fait que ce bassin de contrevenants est similaire à d’autres échantillons de délinquants en ce qui a trait à la nature de leur délinquance, plus particulièrement, au volume, à la variété et à la gravité de leurs crimes. En effet, la majorité des participants rapportent un volume faible de crimes contre la personne et contre les biens alors qu’un petit groupe se démarque par un lambda très élevé (13,1 % des délinquants de l’échantillon sont responsables de 60,3% de tous les crimes rapportés). Environ quatre délinquants sur cinq rapportent avoir commis au moins un crime contre la personne et un crime contre les biens. De plus, plus de 50% de ces derniers rapportent dans au moins quatre sous-catégories. Finalement, bien que les délinquants de notre échantillon aient un IGC (indice de gravité de la criminalité) moyen relativement faible (médiane = 77), près de 40% des contrevenants rapportent avoir commis au moins un des deux crimes les plus graves recensés dans cette étude (décharger une arme et vol qualifié). Le second objectif spécifique était d’explorer, au chapitre quatre, l’interaction entre les caractéristiques personnelles, l’entourage et le style de vie des délinquants dans la formation de situations criminogènes. Les personnes ayant une propension à la délinquance plus élevée semblent avoir tendance à être davantage entourées de personnes criminalisées et à avoir un style de vie plus oisif. L’entourage criminalisé semble également influencer le style de vie de ces délinquants. Ainsi, l’interdépendance entre ces trois éléments facilite la formation plus fréquente de situations criminogènes et crée une conjoncture propice à l’émergence de la délinquance prolifique. Le dernier objectif spécifique de la thèse, qui a été couvert dans le chapitre cinq, était d’analyser l’impact de la formation de situations criminogènes sur la nature de la délinquance. Les analyses de régression linéaires multiples et les arbres de régression ont permis de souligner la contribution des caractéristiques personnelles, de l’entourage et du style de vie dans l’explication de la nature de la délinquance. D’un côté, les analyses de régression (modèles additifs) suggèrent que l’ensemble des éléments favorisant la formation de situations criminogènes apporte une contribution unique à l’explication de la délinquance. D’un autre côté, les arbres de régression nous ont permis de mieux comprendre l’interaction entre les éléments dans l’explication de la délinquance prolifique. En effet, un positionnement plus faible sur certains éléments peut être compensé par un positionnement plus élevé sur d’autres. De plus, l’accumulation d’éléments favorisant la formation de situations criminogènes ne se fait pas de façon linéaire. Ces conclusions sont appuyées sur des proportions de variance expliquée plus élevées que celles des régressions linéaires multiples. En conclusion, mettre l’accent que sur un seul élément (la personne et sa propension à la délinquance ou le contexte et ses opportunités) ou leur combinaison de façon simplement additive ne permet pas de rendre justice à la complexité de l’émergence de la délinquance prolifique. En mettant à l’épreuve empiriquement cette idée généralement admise, cette thèse permet donc de souligner l’importance de considérer l’interaction entre le risque individuel et le risque contextuel dans l’explication de la délinquance prolifique.

O efeito do fator de crescimento de fibroblastos básico aplicado em superfícies radiculares condicionadas com cloridrato de tetraciclina ou EDTA na morfologia e densidade de fibroblastos. Estudo in vitro

Pós-graduação em Odontologia - FOAR

Estudio epidemiológico de las patologías corneales en la especie canina en el Hospital Docente Universitario durante el periodo 2001-2011

Programa de doctorado: Clínica e investigación terapéutica.