Machine learning, neural and statistical classification
Machine learning, neural and statistical classification
Measuring the interestingness of discovered knowledge: A principled approach
Intelligent Data Analysis
Hi-index | 0.00 |
The paper is about how to evaluate the intelligent knowledge. We think that the criteria is not only distinguishing ability and accuracy of the algorithm, but also algorithm robustness, stability and so on. Besides positive characters listed above, a fair evaluation system of intelligent knowledge should also include negative characters, such as storage space, running time, training and testing time and costs and etc. In this paper we look positive and negative characters of data mining algorithm as the outputs and inputs of a decision making unit, and proposed a model to evaluating intelligent knowledge comprehensively using DEA.