José Portela

Robert Gaugl, Eloy Insunza, José Portela et al.

Cross-border electricity exchanges are crucial for operating and planning highly renewable power systems. Many studies reduce spatial granularity to keep the models tractable and prescribe cross-border exchanges exogenously, often by reusing historical import/export time series. Such assumptions become inconsistent as renewable penetration changes the magnitude and timing of flows. This paper proposes a machine-learning (ML) surrogate framework that maps available nodal time series data (e.g., hourly demand and renewable generation) to synthetic, interconnector-level flow time series. The goal is to provide consistent flow profiles that are used as fixed boundary conditions in reduced power system optimization models (PSOMs). To improve downstream feasibility when surrogate flows are imposed in optimization, we further introduce a custom loss for the neural-network surrogate that penalizes physically impossible flow patterns. We demonstrate the framework on a pan-European single-node per country DC optimal power flow setting using the open-source LEGO PSOM with ENTSO-E TYNDP 2024 National Trends assumptions for 2030. We assess two model classes: k-nearest neighbors (KNN) and feedforward neural networks (SQU), using both full and reduced feature sets. The SQU models generalize more robustly than KNN to unseen climate years and substantially improve upon scaled historical benchmarks in terms of predictive accuracy. When imposed as fixed boundary flows in single-node PSOMs, the ML-generated profiles produce outcomes that closely match the results of the full European simulation, while delivering substantial runtime reductions (up to ~500x). These results indicate that ML-based flow surrogates can provide decision-relevant interconnector flows for tractable reduced studies in high-renewable systems.

Roberto Martínez-Cruz, Alvaro J. López-López, José Portela

Transformer-based language models, including ChatGPT, have demonstrated exceptional performance in various natural language generation tasks. However, there has been limited research evaluating ChatGPT's keyphrase generation ability, which involves identifying informative phrases that accurately reflect a document's content. This study seeks to address this gap by comparing ChatGPT's keyphrase generation performance with state-of-the-art models, while also testing its potential as a solution for two significant challenges in the field: domain adaptation and keyphrase generation from long documents. We conducted experiments on six publicly available datasets from scientific articles and news domains, analyzing performance on both short and long documents. Our results show that ChatGPT outperforms current state-of-the-art models in all tested datasets and environments, generating high-quality keyphrases that adapt well to diverse domains and document lengths.

Roberto Martínez-Cruz, Debanjan Mahata, Alvaro J. López-López et al.

In this study, we investigate using graph neural network (GNN) representations to enhance contextualized representations of pre-trained language models (PLMs) for keyphrase extraction from lengthy documents. We show that augmenting a PLM with graph embeddings provides a more comprehensive semantic understanding of words in a document, particularly for long documents. We construct a co-occurrence graph of the text and embed it using a graph convolutional network (GCN) trained on the task of edge prediction. We propose a graph-enhanced sequence tagging architecture that augments contextualized PLM embeddings with graph representations. Evaluating on benchmark datasets, we demonstrate that enhancing PLMs with graph embeddings outperforms state-of-the-art models on long documents, showing significant improvements in F1 scores across all the datasets. Our study highlights the potential of GNN representations as a complementary approach to improve PLM performance for keyphrase extraction from long documents.

Jaime Pizarroso, David Alfaya, José Portela et al.

As Machine Learning models are considered for autonomous decisions with significant social impact, the need for understanding how these models work rises rapidly. Explainable Artificial Intelligence (XAI) aims to provide interpretations for predictions made by Machine Learning models, in order to make the model trustworthy and more transparent for the user. For example, selecting relevant input variables for the problem directly impacts the model's ability to learn and make accurate predictions, so obtaining information about input importance play a crucial role when training the model. One of the main XAI techniques to obtain input variable importance is the sensitivity analysis based on partial derivatives. However, existing literature of this method provide no justification of the aggregation metrics used to retrieved information from the partial derivatives. In this paper, a theoretical framework is proposed to study sensitivities of ML models using metric techniques. From this metric interpretation, a complete family of new quantitative metrics called $α$-curves is extracted. These $α$-curves provide information with greater depth on the importance of the input variables for a machine learning model than existing XAI methods in the literature. We demonstrate the effectiveness of the $α$-curves using synthetic and real datasets, comparing the results against other XAI methods for variable importance and validating the analysis results with the ground truth or literature information.