|
多类在线分类中Bandit信息的价值
The price of bandit information in multiclass online classification |
|
| |
| 课程网址: |
http://videolectures.net/colt2013_daniely_price/
|
| 主讲教师: |
Amit Daniely
|
| 开课单位: |
耶路撒冷希伯来大学
|
| 开课时间: |
2013-08-09
|
| 课程语种: |
英语
|
| 中文简介: |
我们考虑两个场景的多级假设类的在线学习。在完整的信息场景中，学习者将接触到实例及其标签。在强盗场景中，真实的标签并没有被暴露出来，而是表明学习者的预测是否正确。我们表明,错误率在两个场景之间的比例是最多可实现的情况下,和不可知论者的情况。结果是严格到一个对数因子，并基本上回答了一个开放的问题(Daniely et al. - Multiclass learnability and The erm principle)。我们将这些结果应用到一类γ-margin多级线性分类器在Rd。我们表明,该类的强盗错误率可实现的情况下和的不可知论者。这解决了一个开放的问题(Kakade et. al.高效的Bandit算法的在线仿真类预测)。
|
| 课程简介: |
We consider two scenarios of multiclass online learning of a hypothesis class . In the full information scenario, the learner is exposed to instances together with their labels. In the bandit scenario, the true label is not exposed, but rather an indication whether the learner’s prediction is correct or not. We show that the ratio between the error rates in the two scenarios is at most in the realizable case, and in the agnostic case. The results are tight up to a logarithmic factor and essentially answer an open question from (Daniely et. al. - Multiclass learnability and the erm principle). We apply these results to the class of γ-margin multiclass linear classifiers in Rd. We show that the bandit error rate of this class is in the realizable case and in the agnostic case. This resolves an open question from (Kakade et. al. - Efficient bandit algorithms for onlinemulticlass prediction).
|
| 关 键 词: |
在线学习; 实例; 标签; 预测结果; 多类在线分类
|
| 课程来源: |
视频讲座网
|
| 最后编审: |
2019-10-17：cwx
|
| 阅读次数: |
53
|
|
图 书 馆