A Bayesian network approach to the database serch problem in criminal proceedings

Background The 'database search problem', that is, the strengthening of a case - in terms of probative value - against an individual who is found as a result of a database search, has been approached during the last two decades with substantial mathematical analyses, accompanied by lively debate and centrally opposing conclusions. This represents a challenging obstacle in teaching but also hinders a balanced and coherent discussion of the topic within the wider scientific and legal community. This paper revisits and tracks the associated mathematical analyses in terms of Bayesian networks. Their derivation and discussion for capturing probabilistic arguments that explain the database search problem are outlined in detail. The resulting Bayesian networks offer a distinct view on the main debated issues, along with further clarity. Methods As a general framework for representing and analyzing formal arguments in probabilistic reasoning about uncertain target propositions (that is, whether or not a given individual is the source of a crime stain), this paper relies on graphical probability models, in particular, Bayesian networks. This graphical probability modeling approach is used to capture, within a single model, a series of key variables, such as the number of individuals in a database, the size of the population of potential crime stain sources, and the rarity of the corresponding analytical characteristics in a relevant population. Results This paper demonstrates the feasibility of deriving Bayesian network structures for analyzing, representing, and tracking the database search problem. The output of the proposed models can be shown to agree with existing but exclusively formulaic approaches. Conclusions The proposed Bayesian networks allow one to capture and analyze the currently most well-supported but reputedly counter-intuitive and difficult solution to the database search problem in a way that goes beyond the traditional, purely formulaic expressions. The method's graphical environment, along with its computational and probabilistic architectures, represents a rich package that offers analysts and discussants with additional modes of interaction, concise representation, and coherent communication.

New insights on river water chemistry by using non-centred simplicial principal component analysis: a case study

The use of perturbation and power transformation operations permits the investigation of linear processes in the simplex as in a vectorial space. When the investigated geochemical processes can be constrained by the use of well-known starting point, the eigenvectors of the covariance matrix of a non-centred principalcomponent analysis allow to model compositional changes compared with a reference point.The results obtained for the chemistry of water collected in River Arno (central-northern Italy) have open new perspectives for considering relative changes of the analysed variables and to hypothesise the relative effect of different acting physical-chemical processes, thus posing the basis for a quantitative modelling

Compositional ideas in the bayesian analysis of categorical data with application to dose finding clinical trials

Compositional random vectors are fundamental tools in the Bayesian analysis of categorical data.Many of the issues that are discussed with reference to the statistical analysis of compositionaldata have a natural counterpart in the construction of a Bayesian statistical model for categoricaldata.This note builds on the idea of cross-fertilization of the two areas recommended by Aitchison (1986)in his seminal book on compositional data. Particular emphasis is put on the problem of whatparameterization to use

New insights into dendritic cell biology and histiocytosis through a transgenic tumor model

SUMMARY The effective development of an immune response depends on the careful interplay and the regulation between innate and adaptive immunity. As the dendritic cells (DCs) are equipped with many receptors, such as Toll-like receptors, which can detect the presence of infection by recognizing different component of bacteria, fungi and even viruses, they are the among the first cells to respond to the infection. Upon pathogen challenge, the DCs interpret the innate system activation as a maturation signal, resulting in the migration of the DCS to a draining lymph node site. There, activated DCs present efficiently antigens to naïve T cells, which are in turn activated and initiate adaptive immunity. Therefore, DCs are the main connectors between innate and adaptive immune systems. In addition to be the most efficient antigen- presenting cells, DCs play a central role in the regulation of immune responses and immune tolerance. Despite extensive research, many aspects related to DC biology are still unsolved and/or controversial. The low frequency of DCs in vivo often hamper study of DC biology and in vitro-derived DCs are not suited to address certain questions, such as the development of DC. We sought of transforming in vivo the DCs through the specific expression of an oncogene, in order to obtain unlimited numbers of these cells. To achieve this goal, transgenic mouse lines expressing the SV40 Large T oncogene under the control of the CD1 1 c promoter were generated. These transgenic mice are healthy until the age of three to four months without alterations in the DC biology. Thereafter transgenic mice develop a fatal disease that shows features of a human pathology, named histiocytosis, involving DCs. We demonstrate that the disease development in the transgenic mice correlates with a massive accumulation of transformed DCs in the affected organs. Importantly, transformed DCs are immature and fully conserve their capacity to mature in antigen presenting cells. We observe hyperproliferation of transformed DCs only in the sick transgenic mice. Surprisingly, transformed DCs do not proliferate in vitro, but transfer of the transformed DCs into immunodeficient or tolerant host leads to tumor formation. Altoghether, the transgenic mouse lines we have generated represent a valuable tumor model for human histiocytosis, and provide excellent tools to study DC biology. RESUME Le développement d'une réponse immunitaire efficace dépend d'une minutieuse interaction et régulation entre l'immunité innée et adaptative. Comme les cellules dendritiques (DCs) sont équipées de nombreux récepteurs, tels que les récepteurs Toll-like, qui peuvent détecter la présence d'une infection en reconnaissant différents composants bactériens, issus de champignons ou même viraux, elles sont parmi les premières cellules à répondre à l'infection. Suite à la stimulation induite par le pathogène, les DCs interprètent l'activation du système immunitaire inné comme un signal de maturation, résultant dans la migration des DCs vers le ganglion drainant le site d'infection. Là, les DCs actives présentent efficacement des antigènes aux cellules T, qui sont à leur tour activées et initient les systèmes d'immunité adaptative. Ainsi, les DCs forment le lien principal entre les réponses immunitaires innées et adaptatives. En plus d'être les cellules présentatrices d'antigènes les plus efficaces, les DCs jouent un rôle central dans la régulation du système immunitaire et dans le phénomène de tolérance. Malgré des recherches intensives, de nombreux aspects liés à la biologie des DCs sont encore irrésolus et/ou controversés. La faible fréquence des DCs in vivo gêne souvent l'étude de la biologie de ces cellules et les DCs dérivées in vitro ne sont pas adéquates pour adresser certaines questions, telles que le développement des DCs. Afin d'obtenir des quantités illimitées de DCs, nous avons songé à transformer in vivo les DC grâce à l'expression spécifique d'un oncogène. Afin d'atteindre ce but, nous avons généré des lignées de souris transgéniques qui expriment l'oncogène SV40 Large T sous le contrôle du promoter CD1 le. Ces souris transgéniques sont saines jusqu'à l'âge de trois à quatre mois et ne présentent pas d'altération dans la biologie des DCs. Ensuite, les souris transgéniques développent une maladie présentant les traits caractéristiques d'une pathologie humaine nommée histiocytose, qui implique les DCs. Nous montrons que le développement de cette maladie corrèle avec une accumulation massive des DCs transformées dans les organes touchés. De plus, les DCs transformées sont immatures et conservent leur capacité à différencier en cellules présentatrices d'antigène. Nous observons une hyper-prolifération des DCs transformées seulement dans les souris transgéniques malades. Etonnament, les DC transformées ne prolifèrent pas in vitro, par contre, le transfert des DCs transformées dans des hôtes immuno-déficients ou tolérant conduit à la formation de tumeurs. Globalement, les lignées de souris transgéniques que nous avons générées représentent un modèle valide pour l'histiocytose humaine, et de plus, offrent d'excellents outils pour étudier la biologie des DCs.

Uncertainty Quantification Of A Semi-Supervised Support Vector Regression Reservoir Model

Uncertainty quantification of petroleum reservoir models is one of the present challenges, which is usually approached with a wide range of geostatistical tools linked with statistical optimisation or/and inference algorithms. Recent advances in machine learning offer a novel approach to model spatial distribution of petrophysical properties in complex reservoirs alternative to geostatistics. The approach is based of semisupervised learning, which handles both ?labelled? observed data and ?unlabelled? data, which have no measured value but describe prior knowledge and other relevant data in forms of manifolds in the input space where the modelled property is continuous. Proposed semi-supervised Support Vector Regression (SVR) model has demonstrated its capability to represent realistic geological features and describe stochastic variability and non-uniqueness of spatial properties. On the other hand, it is able to capture and preserve key spatial dependencies such as connectivity of high permeability geo-bodies, which is often difficult in contemporary petroleum reservoir studies. Semi-supervised SVR as a data driven algorithm is designed to integrate various kind of conditioning information and learn dependences from it. The semi-supervised SVR model is able to balance signal/noise levels and control the prior belief in available data. In this work, stochastic semi-supervised SVR geomodel is integrated into Bayesian framework to quantify uncertainty of reservoir production with multiple models fitted to past dynamic observations (production history). Multiple history matched models are obtained using stochastic sampling and/or MCMC-based inference algorithms, which evaluate posterior probability distribution. Uncertainty of the model is described by posterior probability of the model parameters that represent key geological properties: spatial correlation size, continuity strength, smoothness/variability of spatial property distribution. The developed approach is illustrated with a fluvial reservoir case. The resulting probabilistic production forecasts are described by uncertainty envelopes. The paper compares the performance of the models with different combinations of unknown parameters and discusses sensitivity issues.

The role of C-Jun N-terminal kinase in a mouse model of intracerebral hemorrhage

Background: Intracerebral hemorrhage (ICH) is a subtype of stroke characterized by a haematoma within the brain parenchyma resulting from blood vessel rupture and with a poor outcome. In ICH, the blood entry into the brain triggers toxicity resulting in a substantial loss of neurons and an inflammatory response. At the same time, blood-brain barrier (BBB) disruption increases water content (edema) leading to growing intracranial pressure, which in turn worsens neurological outcome. Although the clinical presentation is similar in ischemic and hemorrhagic stroke, the treatment is different and the stroke type needs to be determined beforehand by imaging which delays the therapy. C-Jun N-terminal kinases (JNKs) are a family of kinases activated in response to stress stimuli and involved in several pathways such as apoptosis. Specific inhibition of JNK by a TAT-coupled peptide (XG-102) mediates strong neuroprotection in several models of ischemic stroke in rodents. Recently, we have observed that the JNK pathway is also activated in a mouse model of ICH, raising the question of the efficacy of XG-102 in this model. Method: ICH was induced in the mouse by intrastriatal injection of bacterial collagenase (0,1 U). Three hours after surgery, animals received an intravenous injection of 100 mg/kg of XG-102. The neurological outcome was assessed everyday until sacrifice using a score (from 0 to 9) based on 3 behavioral tests performed daily until sacrifice. Then, mice were sacrificed at 6 h, 24 h, 48 h, and 5d after ICH and histological studies performed. Results: The first 24 h after surgery are critical in our ICH mice model, and we have observed that XG-102 significantly improves neurological outcome at this time point (mean score: 1,8 + 1.4 for treated group versus 3,4+ 1.8 for control group, P<0.01). Analysis of the lesion volume revealed a significant decrease of the lesion area in the treated group at 48h (29+ 11mm3 in the treated group versus 39+ 5mm3 in the control group, P=0.04). XG-102 mainly inhibits the edema component of the lesion. Indeed, a significant inhibition Journal of Cerebral Blood Flow & Metabolism (2009) 29, S490-S493 & 2009 ISCBFM All rights reserved 0271-678X/09 $32.00 www.jcbfm.com of the brain swelling was observed in treated animals at 48h (14%+ 13% versus 26+ 9% in the control group, P=0.04) and 5d (_0.3%+ 4.5%versus 5.1+ 3.6%in the control group, P=0.01). Conclusions: Inhibition of the JNK pathway by XG- 102 appears to lead to several beneficial effects. We can show here a significant inhibition of the cerebral edema in the ICH model providing a further beneficial effect of the XG-102 treatment, in addition to the neuroprotection previously described in the ischemic model. This result is of interest because currently, clinical treatment for brain edema is limited. Importantly, the beneficial effects observed with XG-102 in models of both stroke types open the possibility to rapidly treat stroke patients before identifying the stroke subtype by imaging. This will save time which is precious for stroke outcome.

Proposta d'un model predictiu del risc d'ingrés hospitalari amb l'ús d'algoritmes bayesians

Model predictiu basat en xarxes bayesianes que permet identificar els pacients amb major risc d'ingrés a un hospital segons una sèrie d'atributs de dades demogràfiques i clíniques.

Bayesian Markov chain Monte Carlo inversion of geophysical data for the purpose of estimating the hydraulic properties of the unsaturated zone

Ground-penetrating radar (GPR) has the potential to provide valuable information on hydrological properties of the vadose zone because of their strong sensitivity to soil water content. In particular, recent evidence has suggested that the stochastic inversion of crosshole GPR data within a coupled geophysical-hydrological framework may allow for effective estimation of subsurface van-Genuchten-Mualem (VGM) parameters and their corresponding uncertainties. An important and still unresolved issue, however, is how to best integrate GPR data into a stochastic inversion in order to estimate the VGM parameters and their uncertainties, thus improving hydrological predictions. Recognizing the importance of this issue, the aim of the research presented in this thesis was to first introduce a fully Bayesian inversion called Markov-chain-Monte-carlo (MCMC) strategy to perform the stochastic inversion of steady-state GPR data to estimate the VGM parameters and their uncertainties. Within this study, the choice of the prior parameter probability distributions from which potential model configurations are drawn and tested against observed data was also investigated. Analysis of both synthetic and field data collected at the Eggborough (UK) site indicates that the geophysical data alone contain valuable information regarding the VGM parameters. However, significantly better results are obtained when these data are combined with a realistic, informative prior. A subsequent study explore in detail the dynamic infiltration case, specifically to what extent time-lapse ZOP GPR data, collected during a forced infiltration experiment at the Arrenaes field site (Denmark), can help to quantify VGM parameters and their uncertainties using the MCMC inversion strategy. The findings indicate that the stochastic inversion of time-lapse GPR data does indeed allow for a substantial refinement in the inferred posterior VGM parameter distributions. In turn, this significantly improves knowledge of the hydraulic properties, which are required to predict hydraulic behaviour. Finally, another aspect that needed to be addressed involved the comparison of time-lapse GPR data collected under different infiltration conditions (i.e., natural loading and forced infiltration conditions) to estimate the VGM parameters using the MCMC inversion strategy. The results show that for the synthetic example, considering data collected during a forced infiltration test helps to better refine soil hydraulic properties compared to data collected under natural infiltration conditions. When investigating data collected at the Arrenaes field site, further complications arised due to model error and showed the importance of also including a rigorous analysis of the propagation of model error with time and depth when considering time-lapse data. Although the efforts in this thesis were focused on GPR data, the corresponding findings are likely to have general applicability to other types of geophysical data and field environments. Moreover, the obtained results allow to have confidence for future developments in integration of geophysical data with stochastic inversions to improve the characterization of the unsaturated zone but also reveal important issues linked with stochastic inversions, namely model errors, that should definitely be addressed in future research.

Sequential estimation of multipath MIMO-OFDM channels

Wireless “MIMO” systems, employing multiple transmit and receive antennas, promise a significant increase of channel capacity, while orthogonal frequency-division multiplexing (OFDM) is attracting a good deal of attention due to its robustness to multipath fading. Thus, the combination of both techniques is an attractive proposition for radio transmission. The goal of this paper is the description and analysis of a new and novel pilot-aided estimator of multipath block-fading channels. Typical models leading to estimation algorithms assume the number of multipath components and delays to be constant (and often known), while their amplitudes are allowed to vary with time. Our estimator is focused instead on the more realistic assumption that the number of channel taps is also unknown and varies with time following a known probabilistic model. The estimation problem arising from these assumptions is solved using Random-Set Theory (RST), whereby one regards the multipath-channel response as a single set-valued random entity.Within this framework, Bayesian recursive equations determine the evolution with time of the channel estimator. Due to the lack of a closed form for the solution of Bayesian equations, a (Rao–Blackwellized) particle filter (RBPF) implementation ofthe channel estimator is advocated. Since the resulting estimator exhibits a complexity which grows exponentially with the number of multipath components, a simplified version is also introduced. Simulation results describing the performance of our channel estimator demonstrate its effectiveness.

Classificació de sons percussius amb un model generatiu

El treball que presentem a continuació desenvolupa un marc teòric i pràctic per a l'avaluació i estudi d'un model generatiu aplicat a tasques discriminatives de senyals sonores sense component harmònica. El model generatiu està basat en la construcció de l'anomenada deep belief network, un tipus de xarxa neuronal generativa que permet realitzar tasques de classificació i regressió com també de reconstrucció dels seus estats interns.A partir de l'anàlisi realitzada hem pogut obtenir resultats en classificació aparellats amb els resultats de l'estat de l'art de classificadors de sons inharmònics. Tot i no establir una clara superioritat envers altres mètodes, el present treball ha permés desenvolupar una anàlisi per almodel avaluat amb moltes possibilitats de millora en un futur per altres treballs. Al llarg del treball es demostra la seva eficàcia en tasques discriminatives, com també la capacitat de reduir la dimensionalitat de les dades d'entrada al model i les possibilitats de reconstruir els seus estats interns per a obtenir unes sortides de dades de la xarxa similars a les entrades de descriptors.El desenvolupament centrat en la deep belief network ens ha permés construir un entorn unificat d'avaluació de diferents mètodes d'aprenentatge, construcció i adequació de diferents descriptors sonors i una posterior visualització d'estats interns del mateix, que han possibilitat una avaluaciócomparativa i unificada respecte altres mètodes classificadors de l'estat de l'art. També ens ha permés desenvolupar una implementació en un llenguatge d'alt nivell, que ha reportat més significància per a l'enteniment i anàlisi del model avaluat, amb una argumentació més sòlida.Els resultats i l'anàlisi que reportem són significatius i positius per al model avaluat, i degut a la poca literatura existent en el camp de classificació de sons inharmònics com els sons percussius,creiem que és una aportació interessant i significativa per al camp en el que s'engloba el treball.

Presence of an oligodendroglioma-like component in newly diagnosed glioblastoma identifies a pathogenetically heterogeneous subgroup and lacks prognostic value: central pathology review of the EORTC_26981/NCIC_CE.3 trial.

Glioblastoma (GBM) is a morphologically heterogeneous tumor type with a median survival of only 15 months in clinical trial populations. However, survival varies greatly among patients. As part of a central pathology review, we addressed the question if patients with GBM displaying distinct morphologic features respond differently to combined chemo-radiotherapy with temozolomide. Morphologic features were systematically recorded for 360 cases with particular focus on the presence of an oligodendroglioma-like component and respective correlations with outcome and relevant molecular markers. GBM with an oligodendroglioma-like component (GBM-O) represented 15% of all confirmed GBM (52/339) and was not associated with a more favorable outcome. GBM-O encompassed a pathogenetically heterogeneous group, significantly enriched for IDH1 mutations (19 vs. 3%, p = 0.003) and EGFR amplifications (71 vs. 48%, p = 0.04) compared with other GBM, while co-deletion of 1p/19q was found in only one case and the MGMT methylation frequency was alike (47 vs. 46%). Expression profiles classified most of the GBM-O into two subtypes, 36% (5/14 evaluable) as proneural and 43% as classical GBM. The detection of pseudo-palisading necrosis (PPN) was associated with benefit from chemotherapy (p = 0.0002), while no such effect was present in the absence of PPN (p = 0.86). In the adjusted interaction model including clinical prognostic factors and MGMT status, PPN was borderline nonsignificant (p = 0.063). Taken together, recognition of an oligodendroglioma-like component in an otherwise classic GBM identifies a pathogenetically mixed group without prognostic significance. However, the presence of PPN may indicate biological features of clinical relevance for further improvement of therapy.

Not one odour but two: A new model for nestmate recognition.

Recognition systems play a key role in a range of biological processes, including mate choice, immune defence and altruistic behaviour. Social insects provide an excellent model for studying recognition systems because workers need to discriminate between nestmates and non-nestmates, enabling them to direct altruistic behaviour towards closer kin and to repel potential invaders. However, the level of aggression directed towards conspecific intruders can vary enormously, even among workers within the same colony. This is usually attributed to differences in the aggression thresholds of individuals or to workers having different roles within the colony. Recent evidence from the weaver ant Oecophylla smaragdina suggests that this does not tell the whole story. Here I propose a new model for nestmate recognition based on a vector template derived from both the individual's innate odour and the shared colony odour. This model accounts for the recent findings concerning weaver ants, and also provides an alternative explanation for why the level of aggression expressed by a colony decreases as the diversity within the colony increases, even when odour is well-mixed. The model makes additional predictions that are easily tested, and represents a significant advance in our conceptualisation of recognition systems.

Effective continuing professional development for translating shared decision making in primary care: A study protocol.

Background: Shared decision making (SDM) is a process by which a healthcare choice is made jointly by the healthcare professional and the patient. SDM is the essential element of patient-centered care, a core concept of primary care. However, SDM is seldom translated into primary practice. Continuing professional development (CPD) is the principal means by which healthcare professionals continue to gain, improve, and broaden the knowledge and skills required for patient-centered care. Our international collaboration seeks to improve the knowledge base of CPD that targets translating SDM into the clinical practice of primary care in diverse healthcare systems. Methods: Funded by the Canadian Institutes of Health Research (CIHR), our project is to form an international, interdisciplinary research team composed of health services researchers, physicians, nurses, psychologists, dietitians, CPD decision makers and others who will study how CPD causes SDM to be practiced in primary care. We will perform an environmental scan to create an inventory of CPD programs and related activities for translating SDM into clinical practice. These programs will be critically assessed and compared according to their strengths and limitations. We will use the empirical data that results from the environmental scan and the critical appraisal to identify knowledge gaps and generate a research agenda during a two-day workshop to be held in Quebec City. We will ask CPD stakeholders to validate these knowledge gaps and the research agenda. Discussion: This project will analyse existing CPD programs and related activities for translating SDM into the practice of primary care. Because this international collaboration will develop and identify various factors influencing SDM, the project could shed new light on how SDM is implemented in primary care.

Shared knowledge

Agents use their knowledge on the history of the economy in orderto choose what is the optimal action to take at any given moment of time,but each individual observes history with some noise. This paper showsthat the amount of information available on the past evolution of theeconomy is an endogenous variable, and that this leads to overconcentrationof the investment, which can be interpreted as underinvestment in research.It presents a model in which agents have to invest at each period in one of$K$ sectors, each of them paying an exogenous return that follows a welldefined stochastic path. At any moment of time each agent receives an unbiasednoisy signal on the payoff of each sector. The signals differ across agents,but all of them have the same variance, which depends on the aggregate investmentin that particular sector (so that if almost everybody invests in it theperceptions of everybody will be very accurate, but if almost nobody doesthe perceptions of everybody will be very noisy). The degree of hetereogeneityacross agents is then an endogenous variable, evolving across time determining,and being determined by, the amount of information disclosed.As long as both the level of social interaction and the underlying precisionof the observations are relatively large agents behave in a very preciseway. This behavior is unmodified for a huge range of informational parameters,and it is characterized by an excessive concentration of the investment ina few sectors. Additionally the model shows that generalized improvements in thequality of the information that each agent gets may lead to a worse outcomefor all the agents due to the overconcentration of the investment that thisproduces.

Further insights on the French WISC-IV factor structure through Bayesian structural equation modeling

The interpretation of the Wechsler Intelligence Scale for Children-Fourth Edition (WISC-IV) is based on a 4-factor model, which is only partially compatible with the mainstream Cattell-Horn-Carroll (CHC) model of intelligence measurement. The structure of cognitive batteries is frequently analyzed via exploratory factor analysis and/or confirmatory factor analysis. With classical confirmatory factor analysis, almost all crossloadings between latent variables and measures are fixed to zero in order to allow the model to be identified. However, inappropriate zero cross-loadings can contribute to poor model fit, distorted factors, and biased factor correlations; most important, they do not necessarily faithfully reflect theory. To deal with these methodological and theoretical limitations, we used a new statistical approach, Bayesian structural equation modeling (BSEM), among a sample of 249 French-speaking Swiss children (8-12 years). With BSEM, zero-fixed cross-loadings between latent variables and measures are replaced by approximate zeros, based on informative, small-variance priors. Results indicated that a direct hierarchical CHC-based model with 5 factors plus a general intelligence factor better represented the structure of the WISC-IV than did the 4-factor structure and the higher order models. Because a direct hierarchical CHC model was more adequate, it was concluded that the general factor should be considered as a breadth rather than a superordinate factor. Because it was possible for us to estimate the influence of each of the latent variables on the 15 subtest scores, BSEM allowed improvement of the understanding of the structure of intelligence tests and the clinical interpretation of the subtest scores.