Project: Learning with Ordinal Labels: theory, algorithms and applications
- project duration
- 01-DEC-04 – 30-NOV-08
- In this project new machine learning algorithms are developed for classification problems where the data are labeled with an ordinal variable (for example bad, moderate, good). This type of problems occrus when humans participate in the data generation process. This is for example the case in information retrieval, industrial quality control and with data originating from inquiries in medical and social sciences. In this project we will both investigate the theoretical aspects of ordinal regression learning as well as analyze algorithms and some applications.