Abstract:
Classi cation problem is a popular topic within machine learning community.
One of the major problem in classi cation task is how to handle incoming patterns
which are unusual and di erent in some measure. For such novel input the
classi er should be able to distinguish them as unknown type and don't make
any decision. In this work we try to address that problem.
Description:
Dissertation under the supervision of Prof. Nikhil R. Pal, Professor, ECSU,