🧬 Evolutionary Computation

Estimation of distribution algorithms (EDAs) replace the traditional crossover and mutation operators of genetic algorithms with the learning and sampling of a probabilistic model. My PhD dissertation and early research focused on the theoretical and practical foundations of EDAs, including their behavior on different problem types and the analysis of the models they learn.

I am co-author of the monograph Estimation of Distribution Algorithms: A New Tool for Evolutionary Computation (Springer, 2002) and have contributed extensively to the field through publications, tutorials, and special sessions at GECCO, CEC, PPSN, and other conferences.

Research on EDAs for continuous optimization problems, including Gaussian-based models, multivariate probability distributions, and copula-based continuous EDAs. Applications to parametric optimization of geothermal power plants and other engineering problems.

Research on understanding what EDAs learn: analysis of the probabilistic models used by EDAs, the information captured in these models, and how this information relates to the structure of the fitness landscape.

Development of evolutionary algorithms for automatic design of neural network architectures. Research on how to represent, evaluate, and evolve architectures efficiently, including factorized representations that reduce the search space without sacrificing expressiveness.

Research on evolutionary optimization of generative adversarial networks (GANs) and variational autoencoders (VAEs), including the analysis of the interplay between transferable GANs and gradient optimizers.

Use of neuron coverage metrics as objectives to guide neuroevolutionary algorithms for semi-supervised classification. Combining diversity-oriented evolutionary search with neuroscience-inspired coverage criteria.

Research on genetic programming methods for symbolic regression and function learning. Investigation of how GP can discover compact and interpretable mathematical expressions from data.

Contribution to the Handbook of Evolutionary Machine Learning: chapter on EML for Unsupervised Learning, covering evolutionary approaches to clustering, dimensionality reduction, and generative modeling.

Research on evolutionary algorithms for multi-objective optimization, including NSGA-II, MOEA/D, and their applications. Benchmarking MOEAs for continuous multi-objective reinforcement learning problems.

Research (in Basque language) on the exploitation of neuroevolutionary information: learning from the past to build a more effective future. Presented at IkerGazte 2025.

• Santana R, McKay RI and Lozano JA (2013). Symmetry in evolutionary and estimation of distribution algorithms. IEEE TEVC.
• Santana R and Lozano JA (2017). Different scenarios for survival analysis of evolutionary algorithms. Swarm and Evolutionary Computation.
• Garciarena U and Santana R (2016). Evolutionary Optimization of Compiler Flag Selection by Learning and Exploiting Flags Interactions. GECCO 2016.
• Garciarena U, Marti L and Santana R (2021). On the Exploitation of Neuroevolutionary Information: Analyzing the Past for a More Efficient Future. CEC 2021.
• Garciarena U, Marti L and Santana R (2020). Envisioning the Benefits of Back-Drive in Evolutionary Algorithms. CEC 2020.
• Santana R, Mendiburu A and Lozano JA (2016). Evolutionary Approaches to Optimization Problems in Chimera Topologies. GECCO 2016.
• Picek S, Jakobovic D and Santana R (2016). Maximal nonlinearity in balanced boolean functions with even number of inputs, revisited. CEC 2016.
• Larrañaga P, Karshenas H, Bielza C and Santana R (2012). A review on probabilistic graphical models in evolutionary computation. JMLR Workshop.
• Larrañaga P, Karshenas H, Bielza C and Santana R (2013). A Review on Evolutionary Algorithms in Bayesian Network Learning and Inference Tasks. International Journal of Approximate Reasoning.
• Ponce-de-Leon M, Ponce M and Santana R (1996). A genetic algorithm for a Hamiltonian path problem. GECCO 1996.