Complex computer simulations are commonly required for accurate data modelling in many scientific disciplines, making statistical inference challenging due to the intractability of the likelihood evaluation for the observed data. Furthermore, sometimes one is interested on inference drawn over a subset of the generative model parameters while taking into account model uncertainty or misspecification on the remaining nuisance parameters. In this work, we show how non-linear summary statistics can be constructed by minimising inference-motivated losses via stochastic gradient descent such they provided the smallest uncertainty for the parameters of interest. As a use case, the problem of confidence interval estimation for the mixture coefficient in a multi-dimensional two-component mixture model (i.e. signal vs background) is considered, where the proposed technique clearly outperforms summary statistics based on probabilistic classification, which are a commonly used alternative but do not account for the presence of nuisance parameters.Preprint. Work in progress.

New insights into phase separation in colloidal suspensions are provided via a new dynamical theory based on the Polydisperse Lattice-Gas model. The model gives a simplified description of polydisperse colloids, incorporating a hard-core repulsion combined with polydispersity in the strength of the attraction between neighbouring particles. Our mean-field equations describe the local concentration evolution for each of an arbitrary number of species, and for an arbitrary overall composition of the system. We focus on the predictions for the dynamics of colloidal gas-liquid phase separation after a quench into the coexistence region. The critical point and the relevant spinodal curves are determined analytically, with the latter depending only on three moments of the overall composition. The results for the early-time spinodal dynamics show qualitative changes as one crosses a 'quenched' spinodal that excludes fractionation and so allows only density fluctuations at fixed composition. This effect occurs for dense systems, in agreement with a conjecture by Warren that, at high density, fractionation should be generically slow because it requires inter-diffusion of particles. We verify this conclusion by showing that the observed qualitative changes disappear when direct particle-particle swaps are allowed in the dynamics. Finally, the rich behaviour beyond the spinodal regime is examined, where we find that the evaporation of gas bubbles with strongly fractionated interfaces causes long-lived composition heterogeneities in the liquid phase; we introduce a two-dimensional density histogram method that allows such effects to be easily visualized for an arbitrary number of particle species.

No abstract

No abstract

El siguiente artículo analiza la potencialidad para el fomento de la competencia digital que conllevan las apps sobre el patrimonio cultural en el estado español. Para ello, se ha empleado el método OEPE integrando una herramienta elaborada ad hoc para el análisis de estas aplicaciones que cuenta con cinco dimensiones. Para este estudio se ha tomado en cuenta el marco teórico para la competencia digital DigComp propuesta por la Unión Europea y se han extrapolado las variables afines extraídas de la herramienta a este marco. La muestra ha sido de 88 apps seleccionadas a partir de la adaptación de los filtros de inclusión/ exclusión del OEPE para apps de temática patrimonial. Los resultados muestran un impulso bajo en las habilidades más complejas para el desarrollo de la competencia digital, siendo el área del tratamiento de la información la más enriquecedora y la creación de contenidos la menos trabajada. The following article analyses the potential for the development of the Digital Competence of the Spanish apps which subject is the cultural heritage. To carry out this study the OEPE method has been used, integrating an ad hoc tool for the analysis of these applications. For the theoretical framework, it has been taken into account the variables and standards proposed by DigComp’s project done by the European Union and those criteria have been extrapolated to our tool’s dimensions. The sample consisted of 88 apps selected from the adaptation of OEPE’s inclusion/exclusion filters for heritage theme apps. The results show a low impulse in the most complex skills for the development of the Digital Competence, being the area of the information treatment the most enriching and the creation of content the least worked