A classification algorithm based on local cluster centers with a few labeled training examples
Knowledge-Based Systems
| Hi-index | 0.00 |
Because of the complexity of data set with mixed attributes, the traditional clustering algorithms appropriate for this kind of dataset are few and the effect of clustering is not good. K-prototype clustering is one of the most commonly used methods in data mining for this kind of data. We borrow the ideas from the multiple classifiers combing technology, use k-prototype as the basis clustering algorithm to design a multi-level clustering ensemble algorithm in this paper, which adaptively selects attributes for re-clustering. Comparison experiments on Adult data set from UCI machine learning data repository show very competitive results and the proposed method is suitable for data editing.