论文信息 - INSS: A hybrid system for constructive machine learning

INSS: A hybrid system for constructive machine learning

Abstract In this paper we present the INSS system, a new hybrid approach based upon the principles of KBANN networks. It represents an important improvement in comparison with its predecessor because the learning and the knowledge extraction process are faster and are accomplished in an incremental way. INSS offers a new approach applicable to constructive machine learning with high-performance tools, even in the presence of incomplete or erroneous data.

Fernando Santos Osório | Bernard Amy | B. Amy | F. Osório

[1] Olcay Boz,et al. Knowledge Integration and Rule Extraction in Neural Networks , 1995 .

[2] Melanie Hilario,et al. An Overview of Strategies for Neurosymbolic Integration , 1995 .

[3] Sebastian Thrun,et al. The MONK''s Problems-A Performance Comparison of Different Learning Algorithms, CMU-CS-91-197, Sch , 1991 .

[4] Jude W. Shavlik,et al. Knowledge-Based Artificial Neural Networks , 1994, Artif. Intell..

[5] Frédéric Alexandre,et al. Connectionist-Symbolic Integration: From Unified to Hybrid Approaches , 1996 .

[6] William C. Schmidt,et al. A Cascade-Correlation Model of Balance Scale Phenomena , 2000 .

[7] Christian Lebiere,et al. The Cascade-Correlation Learning Architecture , 1989, NIPS.

[8] Fernando Santos Osório,et al. INSS : un système hybride neuro-symbolique pour l'apprentissage automatique constructif , 1998 .

[9] J. Shavlik,et al. The Extraction of Reened Rules from Knowledge-based Neural Networks , 1993 .

[10] Wolfram Schiffmann,et al. Comparison of optimized backpropagation algorithms , 1993, ESANN.

[11] Bruno Orsier,et al. Etude et application de systèmes hybrides neurosymboliques , 1995 .

[12] Arnaud Giacometti. Modèles hybrides de l'expertise , 1992 .

[13] Geoffrey E. Hinton,et al. Learning internal representations by error propagation , 1986 .

[14] Geoffrey G. Towell,et al. Symbolic knowledge and neural networks: insertion, refinement and extraction , 1992 .