Supervised learning by exploring query matrix for support patterns
Yiqiu Han, Wai Pang Lam · 2005
We propose a novel supervised learning framework called SUPE. The learning process SUPE is customized to the instance to be classified called query instance. Given a query instance, the training data is transformed into a query matrix, from which useful patterns are discovered for learning. The final prediction of the class label is performed by combining some statistics of the discovered useful patterns. We show that SUPE conducts the search from specific to general in a significantly reduced hypothesis space. It also facilitates extremely easy training instance maintenance and updates. We have evaluated our method with a real-world problem and benchmark data sets. The results demonstrate that SUPE can achieve good performance and high reliability.