|
结构化输出空间的精确最大边际训练
Accurate Max-margin Training for Structured Output Spaces |
|
| |
| 课程网址: |
http://videolectures.net/icml08_sarawagi_acmmt/
|
| 主讲教师: |
Sunita Sarawagi
|
| 开课单位: |
印度马德拉斯技术学院
|
| 开课时间: |
2008-07-28
|
| 课程语种: |
英语
|
| 中文简介: |
Tsochantaridis等人2005提出了两种用于结构化空间最大边缘训练的公式：边缘缩放和松弛缩放。虽然边缘缩放已被广泛使用，因为它需要与正常结构化预测相同类型的MAP推断，但松弛缩放被认为更准确且表现更好。我们提出了松弛缩放方法的有效变分近似，它解决了其推理瓶颈，同时保留了其比边距缩放的精度优势。我们进一步争辩说，现有的缩放方法并没有在产生违反约束的情况下全面区分真正的标记。我们提出了一个新的最大边际训练器PosLearn，它可以生成违规者，以确保在可分解损失函数的每个位置分离。实际数据集的实证结果表明，PosLearn可以将测试误差降低多达25％。此外，PosLearn违规者可以比松弛违规者更有效地生成;对于许多结构化任务，所需时间只是MAP推理的两倍。
|
| 课程简介: |
Tsochantaridis et al 2005 proposed two formulations for maximum margin training of structured spaces: margin scaling and slack scaling. While margin scaling has been extensively used since it requires the same kind of MAP inference as normal structured prediction, slack scaling is believed to be more accurate and better-behaved. We present an efficient variational approximation to the slack scaling method that solves its inference bottleneck while retaining its accuracy advantage over margin scaling. We further argue that existing scaling approaches do not separate the true labeling comprehensively while generating violating constraints. We propose a new max-margin trainer PosLearn that generates violators to ensure separation at each position of a decomposable loss function. Empirical results on real datasets illustrate that PosLearn can reduce test error by up to 25%. Further, PosLearn violators can be generated more efficiently than slack violators; for many structured tasks the time required is just twice that of MAP inference.
|
| 关 键 词: |
结构化空间; 边缘缩放; 松弛缩放
|
| 课程来源: |
视频讲座网
|
| 最后编审: |
2019-04-21：lxf
|
| 阅读次数: |
74
|
|
图 书 馆