Goal : develop and study new concepts and algorithms of machine learning to solve modeling problems in computational biology and systems biology

Main object of application : complex biological systems, diagnosis, modeling and inference of regulatory networks

Main Research Projects

• Output kernel regression and its extension to semisupervised and transductive tasks
• Module Extraction in autoregressive and state-space models (sparse models, mixture models)
• Incorporation of prior knowledge about dynamics into network inference algorithms
• Biclustering and kernel data : another module extraction method

Main Tools

• Generative approaches : dynamic bayesian networks, HMM
• Ensemble methods : boosting, mixture models, hierarchical mixture of experts
• Kernel-based methods for structured data