Learning hard concepts through constructive induction: framework and rationale
Computational Intelligence
C4.5: programs for machine learning
C4.5: programs for machine learning
Understanding the Crucial Role of AttributeInteraction in Data Mining
Artificial Intelligence Review
Feature Extraction, Construction and Selection: A Data Mining Perspective
Feature Extraction, Construction and Selection: A Data Mining Perspective
Function decomposition in machine learning
Machine Learning and Its Applications
Feature Space Transformation Using Genetic Algorithms
IEEE Intelligent Systems
Computational Methods of Feature Selection (Chapman & Hall/Crc Data Mining and Knowledge Discovery Series)
The Minimum Description Length Principle (Adaptive Computation and Machine Learning)
The Minimum Description Length Principle (Adaptive Computation and Machine Learning)
Fitness Function Comparison for GA-Based Feature Construction
Current Topics in Artificial Intelligence
Data Reduction by Genetic Algorithms and Non-Algebraic Feature Construction: A Case Study
HIS '08 Proceedings of the 2008 8th International Conference on Hybrid Intelligent Systems
Searching for interacting features
IJCAI'07 Proceedings of the 20th international joint conference on Artifical intelligence
Inferring ECA-based rules for ambient intelligence using evolutionary feature extraction
Journal of Ambient Intelligence and Smart Environments
Hi-index | 0.00 |
When used for data reduction, feature selection may successfully identify and discard irrelevant attributes, and yet fail to improve learning accuracy because regularities in the concept are still opaque to the learner. In that case, it is necessary to highlight regularities by constructing new characteristics that abstract the relations among attributes. This paper highlights the importance of feature construction when attribute interaction is the main source of learning difficulty and the underlying target concept is hard to discover by a learner using only primitive attributes. An empirical study centered on predictive accuracy shows that feature construction significantly outperforms feature selection because, even when done perfectly, detection of interacting attributes does not sufficiently facilitates discovering the target concept.