Concrete applications

Numerous concrete issues will be tackled thanks to the results obtained regarding one or more of the research axes:

• Anomaly detection in a multivariable and sequential framework, calculation of risk indicators (e.g. monitoring the operation of aeronautical engines, market risk management in finance, customer risk management in retail banking)
• Design of learning strategies by reinforcement (e.g. customer relationship management through web portals in retail banking)
• Shape/action recognition in a complex environment (e.g. small objects in videos, asymmetric framework)
• Creation of statistical models summarising costly (deterministic) procedures of the simulations of complex systems
• Text-mining in a noisy environment (e.g. Tweets, feedback in aeronautics, Internet)
• Evaluating the performance of rules produced by machine learning algorithms