Analysis of simulation models with fuzzy graph based metamodeling
Performance Evaluation
Principles of data mining
Fast Algorithms for Mining Association Rules in Large Databases
VLDB '94 Proceedings of the 20th International Conference on Very Large Data Bases
Intelligent data analysis
Efficient Algorithms for Mining Closed Itemsets and Their Lattice Structure
IEEE Transactions on Knowledge and Data Engineering
Incremental rule learning based on example nearness from numerical data streams
Proceedings of the 2005 ACM symposium on Applied computing
Top-down induction of decision trees classifiers - a survey
IEEE Transactions on Systems, Man, and Cybernetics, Part C: Applications and Reviews
Hi-index | 0.00 |
The quality of single classification rules for numerical data can be evaluated by different measures. Common measures are the frequency and the confidence of rules beside others. A problem with these measures is that they are valid for a rule only if an uniform distribution of the data, corresponding to the rule, is assumed. Since this is usually not the case, especially when considering high dimensional data, subrules and their properties should be considered additionally. The frequency and the confidence values of the subrules, summarized in a diagram, give more information about the quality of the rules than the properties of the rules solely.