Knowledge-Based Cascade-Correlation

KBCC is an extension of the cascade-correlation algorithm that treats functions encapsulating prior knowledge as black-boxes which, like simple sigmoidal neurons, can be recruited in the network topology. KBCC has been studied on artificial and real tasks and it has successfully reused various kinds of knowledge. This paper surveys the work on KBCC, from old published data to the latest new results. It also describes KBCC’s position in the transfer of knowledge. Likely future research is forecast.

• Type: Conference
• Year: 2000
• Where: IJCNN
• Authors: Thomas R. Shultz, François Rivest
• Keywords: Cascade-correlation Algorithm, IJCNN 2000, Neural Networks, Prior Knowledge, Simple Sigmoidal Neurons