Loading...
Please wait, while we are loading the content...
Similar Documents
KNN Based Document Classifier Using K-d Tree : An Efficient Implementation
| Content Provider | Semantic Scholar |
|---|---|
| Author | Priyanka, T. Swamy, N. N. |
| Copyright Year | 2015 |
| Abstract | The existing techniques for text classification exhibits less efficiency on large collection of documents with higher dimensionality, especially when the corpus contains many noisy or irrelevant term features. To overcome these challenges, in this paper, the strengths of KNN and kd-tree are combined and we implemented the CenKNN classifier in two stages. The dimensionality of documents is reduced by projecting higher ndimensional term feature space to lower ldimensional class-centroid space using centroid classifier in the first stage. Later, the kd-tree search method finds the k nearest neighbours used in KNN classifier. The performance measured by F1-score on the Reuters-21578 dataset, has shown its effectiveness and reduced computation time. |
| File Format | PDF HTM / HTML |
| Alternate Webpage(s) | http://www.ijcscn.com/Documents/Volumes/vol5issue5/ijcscn2015050502.pdf |
| Language | English |
| Access Restriction | Open |
| Content Type | Text |
| Resource Type | Article |
