6 papers
ChronoVAE-HOPE: Beyond Attention -- A Next-Generation VAE Foundation Model for Specialized Time Series Classification
José Alberto Rodríguez, Luis Balderas, Miguel Lastra +2
Time Series Foundation Models (TSFMs) have become a new component of the state-of-the-art in general time series forecasting. However, adapting them to specialized classification t…
KairosHope: A Next-Generation Time-Series Foundation Model for Specialized Classification via Dual-Memory Architecture
Luis Balderas, José Alberto Rodríguez, Miguel Lastra +2
Time Series Foundation Models (TSFMs) have demonstrated notable success in general-purpose forecasting tasks; however, their adaptation to specialized classification problems remai…
MoEITS: A Green AI approach for simplifying MoE-LLMs
Luis Balderas, Miguel Lastra, José M. Benítez
Large language models are transforming all areas of academia and industry, attracting the attention of researchers, professionals, and the general public. In the trek for more powe…
Optimizing Convolutional Neural Network Architecture
Luis Balderas, Miguel Lastra, José M. Benítez
Convolutional Neural Networks (CNN) are widely used to face challenging tasks like speech recognition, natural language processing or computer vision. As CNN architectures get larg…
Can persistent homology whiten Transformer-based black-box models? A case study on BERT compression
Luis Balderas, Miguel Lastra, José M. Benítez
Large Language Models (LLMs) like BERT have gained significant prominence due to their remarkable performance in various natural language processing tasks. However, they come with…
Optimizing Dense Feed-Forward Neural Networks
Luis Balderas, Miguel Lastra, José M. Benítez
Deep learning models have been widely used during the last decade due to their outstanding learning and abstraction capacities. However, one of the main challenges any scientist ha…