Current doctoral theses

Metaheuristic models to the development of intelligent systems for the decision support in software construction
Student: Aurora Ramírez Quesada
Started on: Fall 2013
Keywords: search-based software engineering; software architectures; evolutionary algorithms
Awards and nominations:
  • 2014. Second Prize to the Best Doctoral Thesis proposal. Doctoral Consortium @ JISBD 2014 (Spanish Conference on Software Engineering and Databases).
  • 2014. Best paper of the SBSE Track at GECCO 2014. Nomination to the best conference paper.
  • 2013. Nomination to the Best paper of MAEB 2013 (Spanish Conference on Metaheuristics and Evolutionary Computation).
Abstract: With this PhD thesis, we plan to explore the sinergies between the artificial intelligence and software engineering in order to propose at least the following advances:
  • Exploring new problems in the field of SBSE, mostly focused on the use of metaheuristic techniques to the early software conception and analysis.
  • Developing novel models using different evolutionary techniques and bioinspired approaches.
  • Providing intelligent systems for decision support to software design and construction, mainly focusing on the actual needs of experts engineers and on increasing the interpretability of the resulting models.
  • Developing usable software modules in JCLEC, as well as providing tools in the Software Engineering field to show the applicability and novelty of our proposals. Application of these tools in real environments.
A framework for the analysis and design of evolutionary algorithms: Applications
Student: Juan Ignacio Jaén Mariñas
Started on: Fall 2013
Keywords: evolutionary computation; experimentation framework; JCLEC
Abstract: The key idea of the thesis consists in studying and developing an adaptable, flexible, scalable and visual experimentation framework for Evolutionary Computation (EC). This work emerges from the need to provide graphical and simple support to those practitioners that are not necessarily experts on EC but want to make use of these approaches in their own application domain. One example is the use of EC in data mining. This framework should allow data miners to flexibly adopt different evolutionary approaches to their specific domain. Therefore, a main goal is to minimise the required effort to design, develop and deploy applied experiments on different fields.
See some initial results.

Former PhD students

New challenges in association rule mining: an approach based on genetic programming
PhD student: José María Luna Ariza
Defended on: January 2014
Keywords: genetic programming; data mining; association rule mining
References:
Mining data using automatic programming based on ant colony optimisation
PhD student: Juan Luis Olmo Ortiz
Defended on: March 2013
Keywords: swarm intelligence; ant programming; data mining; classification
References: