Abstract | ||
---|---|---|
Semi-Supervised Support Vector Machines (S3VMs) are an appealing method for us- ing unlabeled data in classiflcation: their ob- jective function favors decision boundaries which do not cut clusters. However their main problem is that the optimization prob- lem is non-convex and has many local min- ima, which often results in suboptimal per- formances. In this paper we propose to use a global optimization technique known as con- tinuation to alleviate this problem. Com- pared to other algorithms minimizing the same objective function, our continuation method often leads to lower test errors. |
Year | DOI | Venue |
---|---|---|
2006 | 10.1145/1143844.1143868 | ICML |
Keywords | Field | DocType |
semi-supervised svms,decision boundary,global optimization technique,optimization problem,semi-supervised support,main problem,objective function,vector machines,appealing method,local minimum,continuation method,support vector machine,global optimization | Continuation method,Mathematical optimization,Global optimization,Computer science,Continuation,Support vector machine,Algorithm,Maxima and minima,Artificial intelligence,Optimization problem,Machine learning | Conference |
ISBN | Citations | PageRank |
1-59593-383-2 | 63 | 3.17 |
References | Authors | |
15 | 3 |
Name | Order | Citations | PageRank |
---|---|---|---|
olivier chapelle | 1 | 5960 | 455.12 |
Mingmin Chi | 2 | 488 | 35.97 |
Alexander Zien | 3 | 1255 | 146.93 |