The Strength of Weak Learnability
Machine Learning
Machine Learning
Automating the analysis and cataloging of sky surveys
Advances in knowledge discovery and data mining
Formal Concept Analysis: Mathematical Foundations
Formal Concept Analysis: Mathematical Foundations
IEEE Transactions on Knowledge and Data Engineering
Improved use of continuous attributes in C4.5
Journal of Artificial Intelligence Research
Distributed data mining: why do more than aggregating models
IJCAI'07 Proceedings of the 20th international joint conference on Artifical intelligence
Scaling up: distributed machine learning with cooperation
AAAI'96 Proceedings of the thirteenth national conference on Artificial intelligence - Volume 1
Review: Formal Concept Analysis in knowledge processing: A survey on models and techniques
Expert Systems with Applications: An International Journal
| Hi-index | 0.00 |
In this work we are interested in the problem of mining very large distributed databases. We propose a distributed data mining technique which produces a meta-classifier that is both predictive and descriptive. This meta-classifier is made of a set of classification rules, which can be refined then validated. The refinement step, proposes to remove from the meta-classifier rules that according to their confidence coefficient, computed by statistical means, would not have a good prediction capability when used with new objects. The validation step uses some samples to fine-tune rules in the rule set resulted from the refinement step. This paper deals especially with the validation process. Indeed, we propose two validation techniques: the first one is very simple and the second one uses a Galois lattice. A detailed description of these processes is presented in the paper, as well as the experimentation proving the viability of our approach.