Learning internal representations by error propagation
Parallel distributed processing: explorations in the microstructure of cognition, vol. 1
Extracting Refined Rules from Knowledge-Based Neural Networks
Machine Learning
Combining Symbolic and Neural Learning
Machine Learning
Neural Networks
Inductive Inference: Theory and Methods
ACM Computing Surveys (CSUR)
Hi-index | 0.00 |
In this paper, we are interested in a hybrid neurosymbolic system. We present the HLS (Hybrid Learning System), a new hybrid approach combining a connexionist module, a symbolic module, a rule extraction module and a rule insertion module. It presents an important improvement in comparison with just a connectionist system. HLS provides a new approach applicable to machine learning with high-performance tools, even in presence of incomplete data. The proposed architecture gives a good performance and allows acquisition/extraction of network knowledge.