Searcharxiv⌕ Search

arXiv subjects

Clara Grazian

Publications and source records attributed to Clara Grazian.

33 records · Page 2Linked to original sources
arXiv2022

Interdisciplinary Discovery of Nanomaterials Based on Convolutional Neural Networks

The material science literature contains up-to-date and comprehensive scientific knowledge of materials. However, their content is unstructured and diverse, resulting in a significant gap in providing sufficient information for material design and synthesis. To this end, we used natural language processing (NLP) and computer vision (CV) techniques based on convolutional neural networks (CNN) to discover valuable experimental-based information about nanomaterials and synthesis methods in energy-material-related publications. Our first system, TextMaster, extracts opinions from texts and classifies them into challenges and opportunities, achieving 94% and 92% accuracy, respectively. Our second system, GraphMaster, realizes data extraction of tables and figures from publications with 98.3\% classification accuracy and 4.3% data extraction mean square error. Our results show that these systems could assess the suitability of materials for a certain application by evaluation of synthesis insights and case analysis with detailed references. This work offers a fresh perspective on mining knowledge from scientific literature, providing a wide swatch to accelerate nanomaterial research through CNN.

cond-mat.mtrl-sci↗
arXiv2022

Bayesian Copula Directional Dependence for causal inference on gene expression data

Modelling and understanding directional gene networks is a major challenge in biology as they play an important role in the architecture and function of genetic systems. Copula Directional Dependence (CDD) can measure the directed connectivity among variables without any strict requirements of distributional and linearity assumptions. Furthermore, copulas can achieve that by isolating the dependence structure of a joint distribution. In this work, a novel extension of the frequentist CDD in the Bayesian setting is introduced. The new method is compared against the frequentist CDD and validated on six gene interactions, three coming from a mouse scRNA-seq dataset and three coming from a bulk epigenome dataset. The results illustrate that the novel proposed Bayesian CDD was able to identify four out of six true interactions with increased robustness compared to the frequentist method. Therefore, the Bayesian CDD can be considered as an alternative way for modeling the information flow in gene networks.

stat.ME↗
arXiv2021

Approximate Bayesian Conditional Copulas

Copula models are flexible tools to represent complex structures of dependence for multivariate random variables. According to Sklar's theorem (Sklar, 1959), any d-dimensional absolutely continuous density can be uniquely represented as the product of the marginal distributions and a copula function which captures the dependence structure among the vector components. In real data applications, the interest of the analyses often lies on specific functionals of the dependence, which quantify aspects of it in a few numerical values. A broad literature exists on such functionals, however extensions to include covariates are still limited. This is mainly due to the lack of unbiased estimators of the copula function, especially when one does not have enough information to select the copula model. Recent advances in computational methodologies and algorithms have allowed inference in the presence of complicated likelihood functions, especially in the Bayesian approach, whose methods, despite being computationally intensive, allow us to better evaluate the uncertainty of the estimates. In this work, we present several Bayesian methods to approximate the posterior distribution of functionals of the dependence, using nonparametric models which avoid the selection of the copula function. These methods are compared in simulation studies and in two realistic applications, from civil engineering and astrophysics.

stat.ME↗
arXiv2019

New formulation of the Logistic-Gaussian process to analyze trajectory tracking data

Improved communication systems, shrinking battery sizes and the price drop of tracking devices have led to an increasing availability of trajectory tracking data. These data are often analyzed to understand animal behavior. In this work, we propose a new model for interpreting the animal movent as a mixture of characteristic patterns, that we interpret as different behaviors. The probability that the animal is behaving according to a specific pattern, at each time instant, is non-parametrically estimated using the Logistic-Gaussian process. Owing to a new formalization and the way we specify the coregionalization matrix of the associated multivariate Gaussian process, our model is invariant with respect to the choice of the reference element and of the ordering of the probability vector components. We fit the model under a Bayesian framework, and show that the Markov chain Monte Carlo algorithm we propose is straightforward to implement. We perform a simulation study with the aim of showing the ability of the estimation procedure to retrieve the model parameters. We also test the performance of the information criterion we used to select the number of behaviors. The model is then applied to a real dataset where a wolf has been observed before and after procreation. The results are easy to interpret, and clear differences emerge in the two phases.

stat.AP↗
arXiv2019

A review of Approximate Bayesian Computation methods via density estimation: inference for simulator-models

This paper provides a review of Approximate Bayesian Computation (ABC) methods for carrying out Bayesian posterior inference, through the lens of density estimation. We describe several recent algorithms and make connection with traditional approaches. We show advantages and limitations of models based on parametric approaches and we then draw attention to developments in machine learning, which we believe have the potential to make ABC scalable to higher dimensions and may be the future direction for research in this area.

stat.CO↗
arXiv2018

On a Loss-based prior for the number of components in mixture models

We propose a prior distribution for the number of components of a finite mixture model. The novelty is that the prior distribution is obtained by considering the loss one would incur if the true value representing the number of components were not considered. The prior has an elegant and easy to implement structure, which allows to naturally include any prior information one may have as well as to opt for a default solution in cases where this information is not available. The performance of the prior, and comparison with existing alternatives, is studied through the analysis of both real and simulated data.

stat.ME↗
arXiv2018

Modelling Preference Data with the Wallenius Distribution

The Wallenius distribution is a generalisation of the Hypergeometric distribution where weights are assigned to balls of different colours. This naturally defines a model for ranking categories which can be used for classification purposes. Since, in general, the resulting likelihood is not analytically available, we adopt an approximate Bayesian computational (ABC) approach for estimating the importance of the categories. We illustrate the performance of the estimation procedure on simulated datasets. Finally, we use the new model for analysing two datasets about movies ratings and Italian academic statisticians' journal preferences. The latter is a novel dataset collected by the authors.

stat.ME↗
arXiv2018

Approximating the Likelihood in Approximate Bayesian Computation

This chapter will appear in the forthcoming Handbook of Approximate Bayesian Computation (2018). The conceptual and methodological framework that underpins approximate Bayesian computation (ABC) is targetted primarily towards problems in which the likelihood is either challenging or missing. ABC uses a simulation-based non-parametric estimate of the likelihood of a summary statistic and assumes that the generation of data from the model is computationally cheap. This chapter reviews two alternative approaches for estimating the intractable likelihood, with the goal of reducing the necessary model simulations to produce an approximate posterior. The first of these is a Bayesian version of the synthetic likelihood (SL), initially developed by Wood (2010), which uses a multivariate normal approximation to the summary statistic likelihood. Using the parametric approximation as opposed to the non-parametric approximation of ABC, it is possible to reduce the number of model simulations required. The second likelihood approximation method we consider in this chapter is based on the empirical likelihood (EL), which is a non-parametric technique and involves maximising a likelihood constructed empirically under a set of moment constraints. Mengersen et al (2013) adapt the EL framework so that it can be used to form an approximate posterior for problems where ABC can be applied, that is, for models with intractable likelihoods. However, unlike ABC and the Bayesian SL (BSL), the Bayesian EL (BCel) approach can be used to completely avoid model simulations in some cases. The BSL and BCel methods are illustrated on models of varying complexity.

stat.CO↗
arXiv2017

Jeffreys priors for mixture estimation: properties and alternatives

While Jeffreys priors usually are well-defined for the parameters of mixtures of distributions, they are not available in closed form. Furthermore, they often are improper priors. Hence, they have never been used to draw inference on the mixture parameters. The implementation and the properties of Jeffreys priors in several mixture settings are studied. It is shown that the associated posterior distributions most often are improper. Nevertheless, the Jeffreys prior for the mixture weights conditionally on the parameters of the mixture components will be shown to have the property of conservativeness with respect to the number of components, in case of overfitted mixture and it can be therefore used as a default priors in this context.

stat.ME↗
arXiv2017

Approximate Bayesian inference in semiparametric copula models

We describe a simple method for making inference on a functional of a multivariate distribution. The method is based on a copula representation of the multivariate distribution and it is based on the properties of an Approximate Bayesian Monte Carlo algorithm, where the proposed values of the functional of interest are weighed in terms of their empirical likelihood. This method is particularly useful when the "true" likelihood function associated with the working model is too costly to evaluate or when the working model is only partially specified.

stat.ME↗
arXiv2015

Jeffreys priors for mixture estimation

While Jeffreys priors usually are well-defined for the parameters of mixtures of distributions, they are not available in closed form. Furthermore, they often are improper priors. Hence, they have never been used to draw inference on the mixture parameters. We study in this paper the implementation and the properties of Jeffreys priors in several mixture settings, show that the associated posterior distributions most often are improper, and then propose a noninformative alternative for the analysis of mixtures.

stat.ME↗
arXiv2015

Accelerating Metropolis-Hastings algorithms by Delayed Acceptance

MCMC algorithms such as Metropolis-Hastings algorithms are slowed down by the computation of complex target distributions as exemplified by huge datasets. We offer in this paper a useful generalisation of the Delayed Acceptance approach, devised to reduce the computational costs of such algorithms by a simple and universal divide-and-conquer strategy. The idea behind the generic acceleration is to divide the acceptance step into several parts, aiming at a major reduction in computing time that out-ranks the corresponding reduction in acceptance probability. Each of the components can be sequentially compared with a uniform variate, the first rejection signalling that the proposed value is considered no further. We develop moreover theoretical bounds for the variance of associated estimators with respect to the variance of the standard Metropolis-Hastings and detail some results on optimal scaling and general optimisation of the procedure. We illustrate those accelerating features on a series of examples

stat.CO↗
arXiv2014

Accelerating Metropolis-Hastings algorithms: Delayed acceptance with prefetching

MCMC algorithms such as Metropolis-Hastings algorithms are slowed down by the computation of complex target distributions as exemplified by huge datasets. We offer in this paper an approach to reduce the computational costs of such algorithms by a simple and universal divide-and-conquer strategy. The idea behind the generic acceleration is to divide the acceptance step into several parts, aiming at a major reduction in computing time that outranks the corresponding reduction in acceptance probability. The division decomposes the "prior x likelihood" term into a product such that some of its components are much cheaper to compute than others. Each of the components can be sequentially compared with a uniform variate, the first rejection signalling that the proposed value is considered no further, This approach can in turn be accelerated as part of a prefetching algorithm taking advantage of the parallel abilities of the computer at hand. We illustrate those accelerating features on a series of toy and realistic examples.

stat.CO↗
arXiv2014

Approximate Integrated Likelihood via ABC methods

We propose a novel use of a recent new computational tool for Bayesian inference, namely the Approximate Bayesian Computation (ABC) methodology. ABC is a way to handle models for which the likelihood function may be intractable or even unavailable and/or too costly to evaluate; in particular, we consider the problem of eliminating the nuisance parameters from a complex statistical model in order to produce a likelihood function depending on the quantity of interest only. Given a proper prior for the entire vector parameter, we propose to approximate the integrated likelihood by the ratio of kernel estimators of the marginal posterior and prior for the quantity of interest. We present several examples.

stat.CO↗