Research Topics

The IASC group have been initially involved in multi-agent systems and soft computing research fields. Innovative methods and algorithms have been proposed in the following contexts:

  1. feature selection/extraction
  2. metrics for performance evaluation of machine learning and pattern recognition systems
  3. hierarchical classification and clustering

Multi-agent systems have been seen from a software engineering perspective, and they have been adopted for fast prototyping of frameworks, systems, and applications.

 

Competences and knowledge

The IASC group in last years acquired knowledge and several competences, including:

 

Topics

  - Soft computing

    • Preprocessiong techniques (feature selection and extraction, correlation  metrics, ROC analysis)
    • Evolutionary computation (in particular, GAs, XCS, ant colony optimization )

 - Classifier ensembles

    • Bagging
    • Boosting
    • Random forestes
    • ECOC

 - Taxonomy handling

    • Hierarchical classification
    • Taxonomy building 

 

Application Fields

  - Bioinformatics

    • Protein secondary  structure prediction

  - Information filtering and retrieval

    • Contextual advertising

primi sui motori con e-max