Markel Irastorza

Applied mathematician with research experience in statistical and predictive modeling, data analysis, mathematical optimization, and machine learning. Experienced in developing and evaluating computational approaches in Python, analysing complex datasets, and applying statistical and machine-learning methods to research problems. Research experience includes predictive modeling of industrial machinery, autoencoder-based instance generation for combinatorial optimization using evolutionary algorithms, and numerical methods for delay systems.

Particularly interested in applied research involving predictive modeling, statistics, optimization, data analysis, and machine learning.

Research Placement Trainee at University of Padova (2026-01 – 2026-07)

Global Training Program

• analyzed engineering and industrial datasets to investigate relationships between machinery characteristics and economic variables.
• Developed and evaluated statistical and machine-learning models for predicting machinery prices from engineering and economic data.
• Investigated model performance and the predictive relationships between machinery attributes and economic variables.
• The project resulted in a scientific paper currently in preparation; my contribution to the data analysis and predictive modeling work has been completed.

Research Intern at BCAM (2024 – 2025)

• Developed a Python-based autoencoder approach for generating novel instances of the Linear Ordering Problem as part of a research project in combinatorial optimization and evolutionary computation.
• Combined machine-learning techniques with optimization algorithms to investigate instance generation and novelty assessment.
• analyzed large collections of combinatorial optimization instances and investigated methods for generating diverse problem instances.
• Applied evolutionary computation and metaheuristic techniques to a research problem at the intersection of machine learning and mathematical optimization.
• The project resulted in the paper “Autoencoder-Based Novelty Search for Instance Generation: A Case Study on the Linear Ordering Problem,” which was presented at the Metaheuristics International Conference (MIC 2026). The paper is forthcoming in Springer Lecture Notes in Computer Science (LNCS) and has not yet been published.

Research Intern at BCAM (2023 – 2024)

• Investigated vibration modeling in manufacturing using delay differential equations.
• Implemented and analyzed numerical methods for approximating eigenvalues and studying stability regions in delay systems.
• Applied spectral analysis, numerical methods, and computational techniques to mathematical modeling problems.
• The project resulted in a research paper published in the Journal of Computational and Applied Mathematics in 2025, “The Tustin Method for Approximating Eigenvalues of Delay Systems.”

MSc in Mathematical Modelization and Investigation, Statistics and Computation – EHU (2024 – 2025)

BSc in Mathematics – EHU (2020 – 2024)