The field of stochastic filtering involves de design, analysis and application to real-world problems of a variety of methodologies for the calibration, tracking and prediction of random dynamical systems, i.e., systems that evolve with time in a stochastic manner, which generally cannot be observed exactly but only contaminated by noise. The single, most widely used stochastic filtering method is the celebrated Kalman filter, which solves the problem of tracking and predicting the evolution of the state variables of a linear and Gaussian state-space stochastic model. The Kalman filter has found a myriad of applications over the years in engineering (automation and control, communications, sensor networks, etc.), geophysics (weather forecasting, climate modeling, etc.), quantitative finance or ecology, to name some relevant fields. When the dynamical models of interest are nonlinear, non-Gaussian, or both, closed form solutions do not exist. Instead, approximation techniques are necessary and a rich field of research has flourished over the last 20 years, producing powerful methodologies such as sigma-point methods for nonlinear Kalman filtering, particle filters, Gaussian mixture-filters and others. Most of these algorithms have quickly gained popularity and found applications in diverse areas of engineering.
This School is related to the applications, methodological advances and basic theory of Stochastic Filtering. We plan to touch upon fundamental (theoretical and methodological) aspects of the field, as well as advanced techniques and recent trends. FASF schedules 5 short courses (one-day tutorials) from renowned international researchers/professors. These tutorials will cover the theoretical foundations of the field, methodological advances and relevant applications. Most courses will include hands-on sessions in the computer lab. The target audience include researchers working on the field, as well as students aiming at being introduced to the topic.