Improved computation of beliefs based on confusion matrix for combining multiple classifiers
2004
One approach among a number of strategies for multiple classifier combination is to calculate the beliefs of confidence based on confusion matrices from individual classifiers. To achieve precise belief computation, based on previous researchers' work, presented is a better algorithm with a more generic capability, showing improved performance.
Keywords:
- Correction
- Source
- Cite
- Save
- Machine Reading By IdeaReader
5
References
23
Citations
NaN
KQI