Learning hard concepts through constructive induction: framework and rationale
Computational Intelligence
The Utility of Knowledge in Inductive Learning
Machine Learning
Using explanation-based and empirical methods in theory revision
Using explanation-based and empirical methods in theory revision
C4.5: programs for machine learning
C4.5: programs for machine learning
AAAI'94 Proceedings of the twelfth national conference on Artificial intelligence (vol. 2)
Overcoming Process Delays with Decision Tree Induction
IEEE Expert: Intelligent Systems and Their Applications
Machine Learning
Making the most of what you've got: using models and data to improve prediction accuracy
Making the most of what you've got: using models and data to improve prediction accuracy
Maximizing Theory Accuracy Through Selective Reinterpretation
Machine Learning
Domain knowledge to support the discovery process: previously discovered knowledge
Handbook of data mining and knowledge discovery
Backward chaining rule induction
Intelligent Data Analysis - Selected papers from IDA2005, Madrid, Spain
Post-analysis of learned rules
AAAI'96 Proceedings of the thirteenth national conference on Artificial intelligence - Volume 1
Searching for meaningful feature interactions with backward-chaining rule induction
IDA'05 Proceedings of the 6th international conference on Advances in Intelligent Data Analysis
Hi-index | 0.00 |
This paper presents a method to incorporate knowledge from possibly imperfect models and domain theories into inductive learning of decision trees for classification. The approach assumes that a model or domain theory reflects useful prior knowledge of the task. Thus the default bias should accept the models predictions as accurate even in the face of somewhat contradictory data which may be unrepresentative or noisy. However our approach allows the system to abandon the model or domain theory, or portions thereof in the fact of sufficiently contradictory data. In particular we use C4.5 to induce decision trees from data that has heen augmented by model or domaintheory-derived features. We weakly bias the system to select model-derived features during decision tree induction but this preference is not dogmatically applied. Our experiments very imperfection in a model the representativeness of data and the veracitv with which model-demed feature are preferred.