Search results for "PROBA"
showing 10 items of 3964 documents
Low-cost scalable discretization, prediction and feature selection for complex systems
2019
The introduced data-driven tool allows simultaneous feature selection, model inference, and marked cost and quality gains.
Defining classifier regions for WSD ensembles using word space features
2006
Based on recent evaluation of word sense disambiguation (WSD) systems [10], disambiguation methods have reached a standstill. In [10] we showed that it is possible to predict the best system for target word using word features and that using this 'optimal ensembling method' more accurate WSD ensembles can be built (3-5% over Senseval state of the art systems with the same amount of possible potential remaining). In the interest of developing if more accurate ensembles, w e here define the strong regions for three popular and effective classifiers used for WSD task (Naive Bayes – NB, Support Vector Machine – SVM, Decision Rules – D) using word features (word grain, amount of positive and neg…
Oxidative stability of virgin olive oil: evaluation and prediction with an adaptive neuro-fuzzy inference system (ANFIS).
2019
Background An adaptive neuro-fuzzy inference system (ANFIS) was employed to predict the oxidative stability of virgin olive oil (VOO) during storage as a function of time, storage temperature, total polyphenol, α-tocopherol, fatty acid profile, ultraviolet (UV) extinction coefficient (K268 ), and diacylglycerols (DAGs). Results The mean total quantities of polyphenols and DAGs were 1.1 and 1.9 times lower in VOOs stored at 25 °C than in the initial samples, and the mean total quantities of polyphenols and DAGs were 1.3 and 2.26 times lower in VOOs stored at 37 °C than in the initial samples, respectively. In a single sample, α-tocopherol was reduced by between 0.52 and 0.91 times during sto…
Improving Speaker-Independent Lipreading with Domain-Adversarial Training
2017
We present a Lipreading system, i.e. a speech recognition system using only visual features, which uses domain-adversarial training for speaker independence. Domain-adversarial training is integrated into the optimization of a lipreader based on a stack of feedforward and LSTM (Long Short-Term Memory) recurrent neural networks, yielding an end-to-end trainable system which only requires a very small number of frames of untranscribed target data to substantially improve the recognition accuracy on the target speaker. On pairs of different source and target speakers, we achieve a relative accuracy improvement of around 40% with only 15 to 20 seconds of untranscribed target speech data. On mul…
Compound conditionals as random quantities and Boolean algebras
2022
Conditionals play a key role in different areas of logic and probabilistic reasoning, and they have been studied and formalised from different angles. In this paper we focus on the de Finetti's notion of conditional as a three-valued object, with betting-based semantics, and its related approach as random quantity as mainly developed by two of the authors. Compound conditionals have been studied in the literature, but not in full generality. In this paper we provide a natural procedure to explicitly attach conditional random quantities to arbitrary compound conditionals that also allows us to compute their previsions. By studying the properties of these random quantities, we show that, in f…
Financial contagion through space-time point processes
2020
AbstractWe propose to study the dynamics of financial contagion by means of a class of point process models employed in the modeling of seismic contagion. The proposal extends network models, recently introduced to model financial contagion, in a space-time point process perspective. The extension helps to improve the assessment of credit risk of an institution, taking into account contagion spillover effects.
Eye-tracking revision processes of translation students and professional translators
2019
Great effort has been made to define and to measure revision competence in translation. However, combined eye tracking and keylogging have hardly been applied in revision research. We believe it is...
A continuous time tug-of-war game for parabolic $p(x,t)$-Laplace type equations
2019
We formulate a stochastic differential game in continuous time that represents the unique viscosity solution to a terminal value problem for a parabolic partial differential equation involving the normalized $p(x,t)$-Laplace operator. Our game is formulated in a way that covers the full range $1<p(x,t)<\infty$. Furthermore, we prove the uniqueness of viscosity solutions to our equation in the whole space under suitable assumptions.
Operational and financial performance of Italian airport companies: A dynamic graphical model
2016
Abstract This paper provides evidence on the relationship within a set of financial and operational indicators for Italian airports over 2008–2014. The limited sample size of national and regional airports suggests to apply the penalised RCON ( V , E ) model, which falls within the class of Gaussian graphical models. It provides both estimate and easy way to visualise conditional independence structures of the variables. Moreover, it is particularly suitable for handling longitudinal data where small number of units and huge number of variables have been collected. Findings highlight that a qualified concept of size matters in determining good financial performance. Specifically, increasing…
Probabilistic cross-validation estimators for Gaussian process regression
2018
Gaussian Processes (GPs) are state-of-the-art tools for regression. Inference of GP hyperparameters is typically done by maximizing the marginal log-likelihood (ML). If the data truly follows the GP model, using the ML approach is optimal and computationally efficient. Unfortunately very often this is not case and suboptimal results are obtained in terms of prediction error. Alternative procedures such as cross-validation (CV) schemes are often employed instead, but they usually incur in high computational costs. We propose a probabilistic version of CV (PCV) based on two different model pieces in order to reduce the dependence on a specific model choice. PCV presents the benefits from both…