Patch Learning for Incremental Classifier Design

Rudy Sicard (1) , Thierry Artières (1)

1 MALIRE - Machine Learning and Information Retrieval

We present a learning algorithm for nominal data. It builds a classifier by adding iteratively a simple patch function that modifies the current classifier. Its main advantage lies in the possibility to learn every patch function parameters optimally from the Bayesian point of view hence avoiding overtraining.

Domaines: Informatique [cs]

Rudy Sicard, Thierry Artières. Patch Learning for Incremental Classifier Design. European Conference on Artificial Intelligence (ECAI 2006), Aug 2006, Riva del Garda, Italy. pp.807-808. ⟨hal-01337594⟩