Loading...
Please wait, while we are loading the content...
Similar Documents
Classification Based On Manifold Semi-Supervised Support Vector Machine
| Content Provider | Semantic Scholar |
|---|---|
| Author | Thanh, Vo Duy Hung, Vo Trung Tuan, Pham Minh Hung, Ho Khac |
| Copyright Year | 2014 |
| Abstract | This article presents a solution along with experimental results for an application of semi-supervised machine learning techniques and improvement on the SVM (Support Vector Machine) based on geodesic model to build text classification applications for Vietnamese language. The objective here is to improve the semi-supervised machine learning by replacing the kernel function of SVM using geodesic distance algorithm. This experiment is implemented on five data layers which are extracted from documents in five topics sports, education, law, international and society news on dantri.com.vn. The experiment compares the results’ accuracies with and without the improved features for SVM semi-supervised machine learning through geodesic distance. This proposed model called manifold semi-supervised machine learning shows significant improvement both in quality and stativity over the pure SVM algorithm. Keywords-text classification, support vector machine (SVM), semi-supervised learning, manifold learning, geodesic model |
| File Format | PDF HTM / HTML |
| Alternate Webpage(s) | http://www.mirlabs.net/ict13/download/paper4.pdf |
| Language | English |
| Access Restriction | Open |
| Content Type | Text |
| Resource Type | Article |
