arXiv subjects

Gabriel Terejanu

Publications and source records attributed to Gabriel Terejanu.

At least 19 recordsLinked to original sources

arXiv2026

Measuring Social Media Polarization Using Large Language Models and Heuristic Rules

Understanding affective polarization in online discourse is crucial for evaluating the societal impact of social media interactions. This study presents a novel framework that leverages large language models (LLMs) and domain-informed heuristics to systematically analyze and quantify affective polarization in discussions on divisive topics such as climate change and gun control. Unlike most prior approaches that relied on sentiment analysis or predefined classifiers, our method integrates LLMs to extract stance, affective tone, and agreement patterns from large-scale social media discussions. We then apply a rule-based scoring system capable of quantifying affective polarization even in small conversations consisting of single interactions, based on stance alignment, emotional content, and interaction dynamics. Our analysis reveals distinct polarization patterns that are event dependent: (i) anticipation-driven polarization, where extreme polarization escalates before well-publicized events, and (ii) reactive polarization, where intense affective polarization spikes immediately after sudden, high-impact events. By combining AI-driven content annotation with domain-informed scoring, our framework offers a scalable and interpretable approach to measuring affective polarization. The source code is publicly available at: https://github.com/hasanjawad001/llm-social-media-polarization.

cs.SI↗

arXiv2025

From Observations to Causations: A GNN-based Probabilistic Prediction Framework for Causal Discovery

Causal discovery from observational data is challenging, especially with large datasets and complex relationships. Traditional methods often struggle with scalability and capturing global structural information. To overcome these limitations, we introduce a novel graph neural network (GNN)-based probabilistic framework that learns a probability distribution over the entire space of causal graphs, unlike methods that output a single deterministic graph. Our framework leverages a GNN that encodes both node and edge attributes into a unified graph representation, enabling the model to learn complex causal structures directly from data. The GNN model is trained on a diverse set of synthetic datasets augmented with statistical and information-theoretic measures, such as mutual information and conditional entropy, capturing both local and global data properties. We frame causal discovery as a supervised learning problem, directly predicting the entire graph structure. Our approach demonstrates superior performance, outperforming both traditional and recent non-GNN-based methods, as well as a GNN-based approach, in terms of accuracy and scalability on synthetic and real-world datasets without further training. This probabilistic framework significantly improves causal structure learning, with broad implications for decision-making and scientific discovery across various fields.

cs.LG↗

arXiv2024

CGLearn: Consistent Gradient-Based Learning for Out-of-Distribution Generalization

Improving generalization and achieving highly predictive, robust machine learning models necessitates learning the underlying causal structure of the variables of interest. A prominent and effective method for this is learning invariant predictors across multiple environments. In this work, we introduce a simple yet powerful approach, CGLearn, which relies on the agreement of gradients across various environments. This agreement serves as a powerful indication of reliable features, while disagreement suggests less reliability due to potential differences in underlying causal mechanisms. Our proposed method demonstrates superior performance compared to state-of-the-art methods in both linear and nonlinear settings across various regression and classification tasks. CGLearn shows robust applicability even in the absence of separate environments by exploiting invariance across different subsamples of observational data. Comprehensive experiments on both synthetic and real-world datasets highlight its effectiveness in diverse scenarios. Our findings underscore the importance of leveraging gradient agreement for learning causal invariance, providing a significant step forward in the field of robust machine learning. The source code of the linear and nonlinear implementation of CGLearn is open-source and available at: https://github.com/hasanjawad001/CGLearn.

cs.LG↗

arXiv2024

Quantifying Influencer Impact on Affective Polarization

In today's digital age, social media platforms play a crucial role in shaping public opinion. This study explores how discussions led by influencers on Twitter, now known as 'X', affect public sentiment and contribute to online polarization. We developed a counterfactual framework to analyze the polarization scores of conversations in scenarios both with and without the presence of an influential figure. Two case studies, centered on the polarizing issues of climate change and gun control, were examined. Our research highlights the significant impact these figures have on public discourse, providing valuable insights into how online discussions can influence societal divisions.

cs.SI↗

arXiv2024

Two-Stage Stance Labeling: User-Hashtag Heuristics with Graph Neural Networks

The high volume and rapid evolution of content on social media present major challenges for studying the stance of social media users. In this work, we develop a two stage stance labeling method that utilizes the user-hashtag bipartite graph and the user-user interaction graph. In the first stage, a simple and efficient heuristic for stance labeling uses the user-hashtag bipartite graph to iteratively update the stance association of user and hashtag nodes via a label propagation mechanism. This set of soft labels is then integrated with the user-user interaction graph to train a graph neural network (GNN) model using semi-supervised learning. We evaluate this method on two large-scale datasets containing tweets related to climate change from June 2021 to June 2022 and gun control from January 2022 to January 2023. Our experiments demonstrate that enriching text-based embeddings of users with network information from the user interaction graph using our semi-supervised GNN method outperforms both classifiers trained on user textual embeddings and zero-shot classification using LLMs such as GPT4. We discuss the need for integrating nuanced understanding from social science with the scalability of computational methods to better understand how polarization on social media occurs for divisive issues such as climate change and gun control.

cs.LG↗

arXiv2023

Evaluation of Induced Expert Knowledge in Causal Structure Learning by NOTEARS

Causal modeling provides us with powerful counterfactual reasoning and interventional mechanism to generate predictions and reason under various what-if scenarios. However, causal discovery using observation data remains a nontrivial task due to unobserved confounding factors, finite sampling, and changes in the data distribution. These can lead to spurious cause-effect relationships. To mitigate these challenges in practice, researchers augment causal learning with known causal relations. The goal of the paper is to study the impact of expert knowledge on causal relations in the form of additional constraints used in the formulation of the nonparametric NOTEARS. We provide a comprehensive set of comparative analyses of biasing the model using different types of knowledge. We found that (i) knowledge that corrects the mistakes of the NOTEARS model can lead to statistically significant improvements, (ii) constraints on active edges have a larger positive impact on causal discovery than inactive edges, and surprisingly, (iii) the induced knowledge does not correct on average more incorrect active and/or inactive edges than expected. We also demonstrate the behavior of the model and the effectiveness of domain knowledge on a real-world dataset.

cs.LG↗

arXiv2023

Machine Fault Classification using Hamiltonian Neural Networks

A new approach is introduced to classify faults in rotating machinery based on the total energy signature estimated from sensor measurements. The overall goal is to go beyond using black-box models and incorporate additional physical constraints that govern the behavior of mechanical systems. Observational data is used to train Hamiltonian neural networks that describe the conserved energy of the system for normal and various abnormal regimes. The estimated total energy function, in the form of the weights of the Hamiltonian neural network, serves as the new feature vector to discriminate between the faults using off-the-shelf classification models. The experimental results are obtained using the MaFaulDa database, where the proposed model yields a promising area under the curve (AUC) of $0.78$ for the binary classification (normal vs abnormal) and $0.84$ for the multi-class problem (normal, and $5$ different abnormal regimes).

cs.LG↗

arXiv2022

From Causal Pairs to Causal Graphs

Causal structure learning from observational data remains a non-trivial task due to various factors such as finite sampling, unobserved confounding factors, and measurement errors. Constraint-based and score-based methods tend to suffer from high computational complexity due to the combinatorial nature of estimating the directed acyclic graph (DAG). Motivated by the `Cause-Effect Pair' NIPS 2013 Workshop on Causality Challenge, in this paper, we take a different approach and generate a probability distribution over all possible graphs informed by the cause-effect pair features proposed in response to the workshop challenge. The goal of the paper is to propose new methods based on this probabilistic information and compare their performance with traditional and state-of-the-art approaches. Our experiments, on both synthetic and real datasets, show that our proposed methods not only have statistically similar or better performances than some traditional approaches but also are computationally faster.

cs.LG↗

arXiv2020

Explainable Deep Modeling of Tabular Data using TableGraphNet

The vast majority of research on explainability focuses on post-explainability rather than explainable modeling. Namely, an explanation model is derived to explain a complex black box model built with the sole purpose of achieving the highest performance possible. In part, this trend might be driven by the misconception that there is a trade-off between explainability and accuracy. Furthermore, the consequential work on Shapely values, grounded in game theory, has also contributed to a new wave of post-explainability research on better approximations for various machine learning models, including deep learning models. We propose a new architecture that inherently produces explainable predictions in the form of additive feature attributions. Our approach learns a graph representation for each record in the dataset. Attribute centric features are then derived from the graph and fed into a contribution deep set model to produce the final predictions. We show that our explainable model attains the same level of performance as black box models. Finally, we provide an augmented model training approach that leverages the missingness property and yields high levels of consistency (as required for the Shapely values) without loss of accuracy.

cs.LG↗

arXiv2019

Approximate Sampling using an Accelerated Metropolis-Hastings based on Bayesian Optimization and Gaussian Processes

Markov Chain Monte Carlo (MCMC) methods have a drawback when working with a target distribution or likelihood function that is computationally expensive to evaluate, specially when working with big data. This paper focuses on Metropolis-Hastings (MH) algorithm for unimodal distributions. Here, an enhanced MH algorithm is proposed that requires less number of expensive function evaluations, has shorter burn-in period, and uses a better proposal distribution. The main innovations include the use of Bayesian optimization to reach the high probability region quickly, emulating the target distribution using Gaussian processes (GP), and using Laplace approximation of the GP to build a proposal distribution that captures the underlying correlation better. The experiments show significant improvement over the regular MH. Statistical comparison between the results from two algorithms is presented.

cs.LG↗

arXiv2019

A Multiple Filter Based Neural Network Approach to the Extrapolation of Adsorption Energies on Metal Surfaces for Catalysis Applications

Computational catalyst discovery involves the development of microkinetic reactor models based on estimated parameters determined from density functional theory (DFT). For complex surface chemistries, the cost of calculating the adsorption energies by DFT for a large number of reaction intermediates can become prohibitive. Here, we have identified appropriate descriptors and machine learning models that can be used to predict part of these adsorption energies given data on the rest of them. Our investigations also included the case when the species data used to train the predictive model is of different size relative to the species the model tries to predict - an extrapolation in the data space which is typically difficult with regular machine learning models. We have developed a neural network based predictive model that combines an established model with the concepts of a convolutional neural network that, when extrapolating, achieves significant improvement over the previous models.

physics.chem-ph↗

arXiv2019

An Approximate Bayesian Long Short-Term Memory Algorithm for Outlier Detection

Long Short-Term Memory networks trained with gradient descent and back-propagation have received great success in various applications. However, point estimation of the weights of the networks is prone to over-fitting problems and lacks important uncertainty information associated with the estimation. However, exact Bayesian neural network methods are intractable and non-applicable for real-world applications. In this study, we propose an approximate estimation of the weights uncertainty using Ensemble Kalman Filter, which is easily scalable to a large number of weights. Furthermore, we optimize the covariance of the noise distribution in the ensemble update step using maximum likelihood estimation. To assess the proposed algorithm, we apply it to outlier detection in five real-world events retrieved from the Twitter platform.

cs.LG↗

arXiv2019

EnLLVM: Ensemble Based Nonlinear Bayesian Filtering Using Linear Latent Variable Models

Real-time nonlinear Bayesian filtering algorithms are overwhelmed by data volume, velocity and increasing complexity of computational models. In this paper, we propose a novel ensemble based nonlinear Bayesian filtering approach which only requires a small number of simulations and can be applied to high-dimensional systems in the presence of intractable likelihood functions. The proposed approach uses linear latent projections to estimate the joint probability distribution between states, parameters, and observables using a mixture of Gaussian components generated by the reconstruction error for each ensemble member. Since it leverages the computational machinery behind linear latent variable models, it can achieve fast implementations without the need to compute high-dimensional sample covariance matrices. The performance of the proposed approach is compared with the performance of ensemble Kalman filter on a high-dimensional Lorenz nonlinear dynamical system.

stat.CO↗

arXiv2017

Environmental Modeling Framework using Stacked Gaussian Processes

A network of independently trained Gaussian processes (StackedGP) is introduced to obtain predictions of quantities of interest with quantified uncertainties. The main applications of the StackedGP framework are to integrate different datasets through model composition, enhance predictions of quantities of interest through a cascade of intermediate predictions, and to propagate uncertainties through emulated dynamical systems driven by uncertain forcing variables. By using analytical first and second-order moments of a Gaussian process with uncertain inputs using squared exponential and polynomial kernels, approximated expectations of quantities of interests that require an arbitrary composition of functions can be obtained. The StackedGP model is extended to any number of layers and nodes per layer, and it provides flexibility in kernel selection for the input nodes. The proposed nonparametric stacked model is validated using synthetic datasets, and its performance in model composition and cascading predictions is measured in two applications using real data.

cs.LG↗

arXiv2017

Approximate Computational Approaches for Bayesian Sensor Placement in High Dimensions

Since the cost of installing and maintaining sensors is usually high, sensor locations are always strategically selected. For those aiming at inferring certain quantities of interest (QoI), it is desirable to explore the dependency between sensor measurements and QoI. One of the most popular metric for the dependency is mutual information which naturally measures how much information about one variable can be obtained given the other. However, computing mutual information is always challenging, and the result is unreliable in high dimension. In this paper, we propose an approach to find an approximate lower bound of mutual information and compute it in a lower dimension. Then, sensors are placed where highest mutual information (lower bound) is achieved and QoI is inferred via Bayes rule given sensor measurements. In addition, Bayesian optimization is introduced to provide a continuous mutual information surface over the domain and thus reduce the number of evaluations. A chemical release accident is simulated where multiple sensors are placed to locate the source of the release. The result shows that the proposed approach is both effective and efficient in inferring QoI.

stat.CO↗

arXiv2014

Model-Driven Data Collection for Biological Systems

For biological experiments aiming at calibrating models with unknown parameters, a good experimental design is crucial, especially for those subject to various constraints, such as financial limitations, time consumption and physical practicability. In this paper, we discuss a sequential experimental design based on information theory for parameter estimation and apply it to two biological systems. Two specific issues are addressed in the proposed applications, namely the determination of the optimal sampling time and the optimal choice of observable. The optimal design, either sampling time or observable, is achieved by an information-theoretic sensitivity analysis. It is shown that this is equivalent with maximizing the mutual information and contrasted with non-adaptive designs, this information theoretic strategy provides the fastest reduction of uncertainty.

stat.AP↗

arXiv2014

Validating Predictions of Unobserved Quantities

The ultimate purpose of most computational models is to make predictions, commonly in support of some decision-making process (e.g., for design or operation of some system). The quantities that need to be predicted (the quantities of interest or QoIs) are generally not experimentally observable before the prediction, since otherwise no prediction would be needed. Assessing the validity of such extrapolative predictions, which is critical to informed decision-making, is challenging. In classical approaches to validation, model outputs for observed quantities are compared to observations to determine if they are consistent. By itself, this consistency only ensures that the model can predict the observed quantities under the conditions of the observations. This limitation dramatically reduces the utility of the validation effort for decision making because it implies nothing about predictions of unobserved QoIs or for scenarios outside of the range of observations. However, there is no agreement in the scientific community today regarding best practices for validation of extrapolative predictions made using computational models. The purpose of this paper is to propose and explore a validation and predictive assessment process that supports extrapolative predictions for models with known sources of error. The process includes stochastic modeling, calibration, validation, and predictive assessment phases where representations of known sources of uncertainty and error are built, informed, and tested. The proposed methodology is applied to an illustrative extrapolation problem involving a misspecified nonlinear oscillator.

physics.data-an↗

arXiv2011

Optimal Data Split Methodology for Model Validation

The decision to incorporate cross-validation into validation processes of mathematical models raises an immediate question - how should one partition the data into calibration and validation sets? We answer this question systematically: we present an algorithm to find the optimal partition of the data subject to certain constraints. While doing this, we address two critical issues: 1) that the model be evaluated with respect to predictions of a given quantity of interest and its ability to reproduce the data, and 2) that the model be highly challenged by the validation set, assuming it is properly informed by the calibration set. This framework also relies on the interaction between the experimentalist and/or modeler, who understand the physical system and the limitations of the model; the decision-maker, who understands and can quantify the cost of model failure; and the computational scientists, who strive to determine if the model satisfies both the modeler's and decision maker's requirements. We also note that our framework is quite general, and may be applied to a wide range of problems. Here, we illustrate it through a specific example involving a data reduction model for an ICCD camera from a shock-tube experiment located at the NASA Ames Research Center (ARC).

physics.data-an↗