Abstract | ||
---|---|---|
The training phase is time-consuming for structured learning, especially for supper-tagging tasks. In this paper, we propose an online distributed Passive-Aggression (PA) by averaging parameters for parallel training, which can reduce the training time significantly. We also give theoretic analysis for its convergence. Experimental results show that our method can accelerate the training process significantly with comparable or even better accuracy. |
Year | DOI | Venue |
---|---|---|
2013 | 10.1007/978-3-642-41491-6_12 | CHINESE COMPUTATIONAL LINGUISTICS AND NATURAL LANGUAGE PROCESSING BASED ON NATURALLY ANNOTATED BIG DATA |
Field | DocType | Volume |
Convergence (routing),Online algorithm,Computer science,Structure learning,Support vector machine,Structured prediction,Artificial intelligence,Machine learning | Conference | 8208 |
Issue | ISSN | Citations |
null | 0302-9743 | 1 |
PageRank | References | Authors |
0.35 | 13 | 4 |
Name | Order | Citations | PageRank |
---|---|---|---|
Jiayi Zhao | 1 | 15 | 2.99 |
Xipeng Qiu | 2 | 556 | 63.33 |
Zhao Liu | 3 | 25 | 10.73 |
Xuanjing Huang | 4 | 1065 | 114.15 |