✖
Manuel Germán Morales
🎯 focusing

1class ContactInformationCard:
2 def __init__(self):
3 self.dept = "Computer Science @ UJA"
4 self.lab = "SiMiDat Research Group"
5 self.email = "mgerman@ujaen.es"
6 self.phone = "+34 953213052"
7
8 def flipCard(self):
9 print("tap on the card to flip.")
10
11 def closeCard(self):
12 print("tap outside to close it.")

Manuel Germán Morales

mgermanm0.github.io

Manuel Germán Morales /manˈwel xeɾˈman moˈɾales/

I am a Predoctoral Researcher and PhD student in the Information and Communication Technologies program at the Intelligent Systems and Data Mining (SiMiDat) Research Group, affiliated with the Universidad de Jaén (UJA) and the DaSCI Institute. 🎯

I hold both a Bachelor’s and a Master’s degree in Computer Engineering from UJA. My current doctoral research focuses on advancing Deep Learning architectures, specifically exploring Time Series Forecasting and the adaptation of Foundation Models.

My research focuses on the intersection of Time Series Analysis and Foundation Models. I work on developing efficient adaptation techniques to leverage large-scale pre-trained models for high-precision forecasting and intelligent data mining.

Machine Learning: Foundation Models, Transfer Learning, and Parameter-Efficient Fine-Tuning (PEFT)
Data Science: Advanced Time Series Forecasting, Evolutionary Computing and Metaheuristics

📜 updates


📚 publications

A Didactic Overview of Transformer Applications: Model Variations and User Guidelines
Logic Journal of the IGPL (Oxford University Press) [Published (2026)]
MI Cabrera-Bermejo, M. Germán, D. de la Rosa, MJ del Jesus, AJ Rivera, D Elizondo, F Charte, MD Pérez Godoy.
[PDF]
Abstract: A comprehensive guide on Transformer architectures and their practical applications. This work provides user guidelines and a structural overview of model variations, serving as a didactic resource for understanding the deployment of Transformer models in various domains. ... See More
Transfer Learning with Foundational Models for Time Series Forecasting using Low-Rank Adaptations
Information Fusion (Elsevier) [Published (2025)]
M. Germán-Morales, A.J. Rivera-Rivas, M.J. del Jesus, C.J. Carmona
[PDF]
Abstract: Leveraging Low-Rank Adaptation (LoRA) to adapt Foundation Models for time series. We propose a methodology to bridge the gap between Large Language Models and time series forecasting, achieving high-performance results with significantly reduced trainable parameters. ... See More
Reducing Experimentation Training Overhead in Time Series Forecasting through LLM Adaptation
Int. Conf. on Intelligent Data Engineering and Automated Learning (IDEAL) [Published (2025)]
M. Germán-Morales, A.J. Rivera, M.J. del Jesus, C.J. Carmona
[PDF]
Abstract: Addressing the computational costs of training deep learning models for forecasting. This study demonstrates how adapting Large Language Models (LLMs) can minimize experimental overhead while maintaining competitive forecasting accuracy. ... See More


👨‍🏫 teaching

Programming and Network Administration
Lab Instructor | 2024-2025
University of Jaen (UJA)
BSc in Computer Engineering, lab sessions.
Programming and Network Administration
Lab Instructor | 2025-2026
University of Jaen (UJA)
BSc in Computer Engineering, lab sessions.
Data Science and Learning
Teaching Assistant | 2024-2025
Spanish Association for Artificial Intelligence (AEPIA)
MSc in AI Research.
Data Science and Learning
Teaching Assistant | 2025-2026
Spanish Association for Artificial Intelligence (AEPIA)
MSc in AI Research.

‎