I work as a scientist and lecturer at the Department of Computing Science at Umeå University, in Sweden. My room is at the fourth floor in the MIT building, and the room number is B 421. I can be reached by phone on number 090 - 16 63 98 (i.e. +46 - 90 16 63 98 from abroad) or by email: marten@cs.umu.se.
At the department, and with the help of Anders Sjö at the department of Computer Science in Lund, we have developed material for a first course in scientific computing with emphasis on numerical analysis (only in swedish so far). If you want a copy of the material send me an email. The material is always under reconstruction and all comments and suggestions are appreciated. The second more advanced part will hopefully appear soon.
At the department we have a small group focusing on inverse problems, and especially nonlinear least squares problems with different applications. We are currently developing methods for solving ill-posed problems where the solution is not well defined. Such ill-posed problems arise in many applications such as artificial neural networks, image reconstruction and total least squares. A central tool is the nonlinear L-curve and its dual a-curve. For more details see the publications.
Another project I am involved in is parameter estimation in differential equations. The main ideas are to use weighted nonlinear least squares to minimize the difference between the model given by a differential equation and measured data. One specially interesting application is when the differential equation is a time dependent partial differential equation with diffusion and convection. These kind of parameter estimation problems appear in several different areas in the pulp industry.
If you are interested to read more about my work, you can have a look at my list of publications. I'll be happy to send you any article you want if you just drop me an email, including the name of the report and your postal address.
Mårten Gulliksson, Department of Computing Science
