I am studying Pattern Recognition. How is it possible that humans are so very good in recognizing things, and a computer is not? My main research interest is one-class classification (with my toolbox dd_tools). but to be honest, I like any learning method which does not include tons of 'magic' parameters. When you have a look at my publications, you may get an impression.
I am not only working on dd_tools, I also have
- an overview of the functions in dd_tools,
- an overview of one-class datasets.
- a multiple instance learning (MIL) toolbox,
- a results toolbox,
- other software.
In addition, Multiple Instance datasets can be downloaded from MIproblems.org.