Sub navigation

• Seminars and Lectures
• Meetings and Conferences
• Other Events
• Visitor Information

Breadcrumb

Modern challenges of curve modelling: inverse problems and qualitative constraints

Contact Person: Azita Ghassemi

• E-mail for more information
• Meeting website

Organisers: Aurore Delaige, Arne Kovac, Azita Ghassemi and Peter Green

Description
The scope of the workshop will be inverse problems, and curve estimation under qualitative constraints.

Within inverse problems, there will be a particular focus in this workshop on the problem of measurement error, where the goal is to develop statistical methods taking this into account. Measurement error arises in many applications but has recently been most important in analysing medical and econometrics data.

Qualitative assumptions have successfully been used in improving the performance of existing regression methodology and there is clear potential for further improvement. Constraints such as monotonicity or bounds on the number of local extrema are easy to interpret and can be used to replace quantitative assumptions such as smoothness which are often difficult to justify.