arXiv subjects

Pablo González

Publications and source records attributed to Pablo González.

At least 19 recordsLinked to original sources

arXiv2026

Geometry-Aware Bayesian Quantification via Compositional Data Analysis

Accurately estimating the unknown target label distribution is the critical first step for adapting to label shift. This task, widely known as quantification or class prevalence estimation, has recently seen significant advances through continuous KDE-based methods which model the density of multiclass classifier posteriors. Posterior vectors might be regarded as compositional data, since they lie on the probability simplex. However, existing KDE-based quantifiers typically rely on Euclidean Gaussian kernels, which ignore simplex geometry and incorrectly assign probability mass outside its boundaries. We introduce a geometry-aware KDE model for multiclass quantification based on log-ratio representations and Aitchison geometry, together with a shrinkage regularization that improves robustness near the simplex boundary. Combined with a maximum-likelihood interpretation of KDE-based quantification, we derive both point-estimation and Bayesian inference procedures for class prevalences. Experiments on 42 datasets across tabular, text, and image domains show that the proposed method is competitive with state-of-the-art quantifiers, often improving over standard KDE-based baselines, while also yielding strong results among Bayesian quantification methods.

cs.LG↗

arXiv2026

A Computational Framework for Cross-Domain Mission Design and Onboard Cognitive Decision Support

The design of distributed autonomous systems for operation beyond reliable ground contact presents a fundamental tension: as round-trip communication latency grows, the set of decisions delegable to ground operators shrinks. This paper establishes a unified computational methodology for quantifying and comparing this constraint across seven heterogeneous mission architectures, spanning Earth low-orbit surveillance constellations, Mars orbital navigation systems, autonomous underwater mine-clearing swarms, deep-space inter-satellite link networks, and outer-planet in-situ buoy platforms. We introduce the Autonomy Necessity Score, a log-domain latency metric mapping each system continuously from the ground-dependent to the fully-autonomous regime, grounded in nine independently validated computational studies covering Walker spherical-cap coverage mechanics, infrared Neyman-Pearson detection, Extended Kalman Filter hypersonic tracking, cross-mission RF and acoustic link budgets spanning seven orders of magnitude in range, Monte Carlo science-yield sensitivity for TDMA inter-satellite protocols, cross-architecture power budget sizing, distributed magnetic-signature formation emulation, and Arrhenius-corrected cryogenic swarm reliability. Building on this foundation, we evaluate an LLM-based Autonomous Mission Decision Support layer in which three foundation models (Llama-3.3-70B, DeepSeek-V3, and Qwen3-A22B) are queried live via the Nebius AI Studio API across ten structured anomaly scenarios derived directly from the preceding analyses. The best-performing model achieves 80% decision accuracy against physics-grounded ground truth, with all 180 inference calls completing within a 2 s latency budget consistent with radiation-hardened edge deployment, establishing the viability of foundation models as an onboard cognitive layer for high-ANS missions.

eess.SY↗

arXiv2025

Quantification via Gaussian Latent Space Representations

Quantification, or prevalence estimation, is the task of predicting the prevalence of each class within an unknown bag of examples. Most existing quantification methods in the literature rely on prior probability shift assumptions to create a quantification model that uses the predictions of an underlying classifier to make optimal prevalence estimates. In this work, we present an end-to-end neural network that uses Gaussian distributions in latent spaces to obtain invariant representations of bags of examples. This approach addresses the quantification problem using deep learning, enabling the optimization of specific loss functions relevant to the problem and avoiding the need for an intermediate classifier, tackling the quantification problem as a direct optimization problem. Our method achieves state-of-the-art results, both against traditional quantification methods and other deep learning approaches for quantification. The code needed to reproduce all our experiments is publicly available at https://github.com/AICGijon/gmnet.

cs.LG↗

arXiv2024

Quantification using Permutation-Invariant Networks based on Histograms

Quantification, also known as class prevalence estimation, is the supervised learning task in which a model is trained to predict the prevalence of each class in a given bag of examples. This paper investigates the application of deep neural networks to tasks of quantification in scenarios where it is possible to apply a symmetric supervised approach that eliminates the need for classification as an intermediary step, directly addressing the quantification problem. Additionally, it discusses existing permutation-invariant layers designed for set processing and assesses their suitability for quantification. In light of our analysis, we propose HistNetQ, a novel neural architecture that relies on a permutation-invariant representation based on histograms that is specially suited for quantification problems. Our experiments carried out in the only quantification competition held to date, show that HistNetQ outperforms other deep neural architectures devised for set processing, as well as the state-of-the-art quantification methods. Furthermore, HistNetQ offers two significant advantages over traditional quantification methods: i) it does not require the labels of the training examples but only the prevalence values of a collection of training bags, making it applicable to new scenarios; and ii) it is able to optimize any custom quantification-oriented loss function.

cs.LG↗

arXiv2024

Kernel Density Estimation for Multiclass Quantification

Several disciplines, like the social sciences, epidemiology, sentiment analysis, or market research, are interested in knowing the distribution of the classes in a population rather than the individual labels of the members thereof. Quantification is the supervised machine learning task concerned with obtaining accurate predictors of class prevalence, and to do so particularly in the presence of label shift. The distribution-matching (DM) approaches represent one of the most important families among the quantification methods that have been proposed in the literature so far. Current DM approaches model the involved populations by means of histograms of posterior probabilities. In this paper, we argue that their application to the multiclass setting is suboptimal since the histograms become class-specific, thus missing the opportunity to model inter-class information that may exist in the data. We propose a new representation mechanism based on multivariate densities that we model via kernel density estimation (KDE). The experiments we have carried out show our method, dubbed KDEy, yields superior quantification performance with respect to previous DM approaches. We also investigate the KDE-based representation within the maximum likelihood framework and show KDEy often shows superior performance with respect to the expectation-maximization method for quantification, arguably the strongest contender in the quantification arena to date.

cs.LG↗

arXiv2023

Binary Quantification and Dataset Shift: An Experimental Investigation

Quantification is the supervised learning task that consists of training predictors of the class prevalence values of sets of unlabelled data, and is of special interest when the labelled data on which the predictor has been trained and the unlabelled data are not IID, i.e., suffer from dataset shift. To date, quantification methods have mostly been tested only on a special case of dataset shift, i.e., prior probability shift; the relationship between quantification and other types of dataset shift remains, by and large, unexplored. In this work we carry out an experimental analysis of how current quantification algorithms behave under different types of dataset shift, in order to identify limitations of current approaches and hopefully pave the way for the development of more broadly applicable methods. We do this by proposing a fine-grained taxonomy of types of dataset shift, by establishing protocols for the generation of datasets affected by these types of shift, and by testing existing quantification methods on the datasets thus generated. One finding that results from this investigation is that many existing quantification methods that had been found robust to prior probability shift are not necessarily robust to other types of dataset shift. A second finding is that no existing quantification method seems to be robust enough to dealing with all the types of dataset shift we simulate in our experiments. The code needed to reproduce all our experiments is publicly available at https://github.com/pglez82/quant_datasetshift.

cs.LG↗

arXiv2022

White Paper on Forward Physics, BFKL, Saturation Physics and Diffraction

The goal of this whitepaper is to give a comprehensive overview of the rich field of forward physics. We discuss the occurrences of BFKL resummation effects in special final states, such as Mueller-Navelet jets, jet gap jets, and heavy quarkonium production. It further addresses TMD factorization at low x and the manifestation of a semi-hard saturation scale in (generalized) TMD PDFs. More theoretical aspects of low x physics, probes of the quark gluon plasma, as well as the possibility to use photon-hadron collisions at the LHC to constrain hadronic structure at low x, and the resulting complementarity between LHC and the EIC are also presented. We also briefly discuss diffraction at colliders as well as the possibility to explore further the electroweak theory in central exclusive events using the LHC as a photon-photon collider.

hep-ph↗

arXiv2020

Topological Equivalence of nonautonomous difference equations with a family of dichotomies on the half line

A linear system of difference equations and a nonlinear perturbation are considered, we obtain sufficient conditions to ensure the topological equivalence between them, namely, the linear part satisfies a property of dichotomy on the positive half--line while the nonlinearity has some boundedness and Lipschitzness conditions. As a consequence, we study the asymptotical stability and its preservation by topological equivalence.

math.CA↗

arXiv2019

Generalized Galileon Scenario Inspires Chaotic Inflation

We study chaotic inflation with a Galileon-like self interaction $G(ϕ,X)\Box ϕ$, where $G(ϕ,X)\propto X^{n}$. General conditions required for successful inflation are deduced and discussed from the background and cosmological perturbations under slow-roll approximation. Interestingly, it is found that in the regime where the Galileon term dominates over the standard kinetic term, the tensor-to-scalar ratio becomes significantly suppressed in comparison to the standard expression in General Relativity (GR). Particularly, we find the allowed range in the space of parameters characterizing the chaotic quadratic and quartic inflation models by considering the current observational data of Planck from the $n_{\mathcal{S}}-r$ plane. Finally, we discuss about the issue if the Galileon term is dominant by the end of inflation, this can affect the field oscillation during reheating.

gr-qc↗

arXiv2018

Covariant evolution of perturbations during reheating in two-field inflation

We develop a covariant method for studying the effects of a reheating phase on the primordial adiabatic and isocurvature perturbations in two-field models of inflation. To model the decay of the scalar fields into radiation at the end of inflation, we introduce a prescription in which radiation is treated as an additional effective scalar field, requiring us to extend the two-field setup into a three-field system. In this prescription, the coupling between radiation and the scalars can be interpreted covariantly in terms geometrical quantities that parametrize the evolution of a background trajectory in a three-field space. In order to obtain concrete results, we consider two scenarios characterized for having unsuppressed isocurvature fluctuations at the end of inflation: (1) canonical two-field inflation with the product exponential potential, which sources a large negative amount of non-gaussianity and, (2) two-field inflation with an ultra-light field, a model in which the isocurvature mode becomes approximately massless, and its interaction with the curvature perturbation persists during the entire period of inflation. In both cases we discuss how their predictions are modified by the coupling of the scalar fields to the radiation fluid.

hep-th↗

arXiv2017

$\tildeδ$ Gravity and Schwarzschild Solution

$\tildeδ$ Gravity is a gravitational field model, where the geometry is governed by two symmetric tensors, $g_{μν}$ and $\tilde{g}_{μν}$, and new matter fields ($\tildeδ$ Matter fields) are added to the original matter fields. These new components appear motivated by a new symmetry, called $\tildeδ$ symmetry. In previous works, the model is used to explain the expansion of the Universe without Dark Energy and use $\tildeδ$ Matter as a source of Dark Matter. In this paper, we will developed an initial study of Schwarzschild geometry in $\tildeδ$ Gravity to complement the Dark Matter analysis and introduce other phenomena. We will get a modified deflection of light produced by the sun, the perihelion precession and black hole solution.

gr-qc↗

arXiv2017

Non-Relativistic $\tildeδ$ Gravity: A Description of Dark Matter

$\tildeδ$ Gravity is a gravitational field model, where the geometry is governed by two symmetric tensors, $g_{μν}$ and $\tilde{g}_{μν}$, and new matter fields ($\tildeδ$ Matter fields) are added to the original matter fields. These new components appear motivated by a new symmetry, called $\tildeδ$ symmetry. In previous works, the model is used to explain the expansion of the Universe without Dark Energy. This result and the additional contribution to the mass by $\tildeδ$ Matter are motivations to study the Dark Matter effect with $\tildeδ$ Gravity. In this work, we will derive the Non-Relativistic limit to obtain a correction to the rotation velocity in a galaxy, then we will analyze the most common galaxy density profiles to describe Dark Matter.

gr-qc↗

arXiv2017

$\tildeδ$ Gravity, $\tildeδ$ matter and the accelerated expansion of the Universe

A gravitational field model based on two symmetric tensors, $g_{μν}$ and $\tilde{g}_{μν}$, is presented. In this model, new matter fields are added to the original matter fields, motivated by an additional symmetry ($\tildeδ$ symmetry). We call them $\tildeδ$ matter fields. We find that massive particles do not follow geodesics, while trajectories of massless particles are null geodesics of an effective metric. Then we study the Cosmological case, where we get an accelerated expansion of the Universe without dark energy.

gr-qc↗

arXiv2016

Device-Independent Certification of a Nonprojective Qubit Measurement

Quantum measurements on a two-level system can have more than two independent outcomes, and in this case, the measurement cannot be projective. Measurements of this general type are essential to an operational approach to quantum theory, but so far, the nonprojective character of a measurement can only be verified experimentally by already assuming a specific quantum model of parts of the experimental setup. Here, we overcome this restriction by using a device-independent approach. In an experiment on pairs of polarization-entangled photonic qubits we violate by more than 8 standard deviations a Bell-like correlation inequality that is valid for all sets of two-outcome measurements in any dimension. We combine this with a device-independent verification that the system is best described by two qubits, which therefore constitutes the first device-independent certification of a nonprojective quantum measurement.

quant-ph↗

arXiv2015

Electroweak standard model with very special relativity

The Very Special Relativity Electroweak Standard Model (VSR EW SM) is a theory with $SU(2)_L \times U(1)_R$ symmetry, with the same number of leptons and gauge fields as in the usual Weinberg-Salam (WS) model. No new particles are introduced. The model is renormalizable and unitarity is preserved. However, photons obtain mass and the massive bosons obtain different masses for different polarizations. Besides, neutrino masses are generated. A VSR invariant term will produce neutrino oscillations and new processes are allowed. In particular, we compute the rate of the decays $μ\rightarrow e + γ$. All these processes, which are forbidden in the Electroweak Standard Model, put stringent bounds on the parameters of our model and measure the violation of Lorentz invariance. We investigate the canonical quantization of this non-local model. Second quantization is carried out obtaining a well defined particle concept. Additionally, we do a counting of the degrees of freedom associated to the gauge bosons involved in this work, after Spontaneous Symmetry Breaking has been realized. Violations of Lorentz invariance have been predicted by several theories of Quantum Gravity \cite{amu}. It is a remarkable possibility that the low energy effects of Lorentz violation induced by Quantum Gravity could be contained in the non-local terms of the VSR EW SM.

hep-ph↗

arXiv2012

Cosmology in Delta-Gravity

We present a model of the gravitational field based on two symmetric tensors. Gravity is affected by the new field, but outside matter the predictions of the model coincide exactly with general relativity, so all classical tests are satisfied. We find that massive particles do not follow a geodesic while massless particles trajectories are null geodesics of an effective metric. We study the Cosmological case, where we get an accelerated expansion of the universe without dark energy. We also introduce the possibility to explain dark matter with $\tildeδ$ gravity.

gr-qc↗

arXiv2011

A Finite Quantum Gravity Field Theory Model

We discuss the quantization of Delta gravity, a two symmetric tensors model of gravity. This model, in Cosmology, shows accelerated expansion without a cosmological constant. We present the $\tildeδ$ transformation which defines the geometry of the model. Then we show that all delta type models live at one loop only. We apply this to General Relativity and we calculate the one loop divergent part of the Effective Action showing its null contribution in vacuum, implying a finite model. Then we proceed to study the existence of ghosts in the model. Finally, we study the form of the finite quantum corrections to the classical action of the model.

gr-qc↗

arXiv2009

Lorentz Violation on The Primordial Baryogenesis

Recently many studies have considered the possibility of a Lorentz Invariance Violation (LIV), and explored its consequences in a wide range of experiments. If this is true, a LIV could explains some mysteries in Cosmology. In this paper specifically, we will analyze the effects on The Primordial Baryogenesis because it is one of the more important and mysterious phenomena of the Big-Bang, that happened at very high energies, so we have a real chance to obtain an important effect. We will see that this effect could exist, depending directly on the temperature, that is very high at this time in the history of the Universe. So, it is possible to use this result as a test for a LIV and explore the possibility that the boson that started the baryogenesis explains, in part, the dark matter. We will obtain estimates about the beginning time of the baryogenesis and the boson mass too, that come directly from the LIV.

hep-ph↗