arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2103.00241cs.LGcs.CV
收藏

基于迁移学习的自动机器学习

Improved Automated Machine Learning from Transfer Learning

Duke University(杜克大学)

Cat P. Le, Mohammadreza Soltani, Robert Ravier, Vahid Tarokh

摘要:

在本文中,我们提出了一个神经架构搜索框架的基础上,一些基线任务和目标任务之间的相似性度量。我们首先定义的概念的任务相似性的基础上的Fisher信息矩阵的对数行列式。接下来,我们计算每个基线任务与目标任务的任务相似度。通过利用目标和一组学习的基线任务之间的关系,可以显著地减小用于目标任务的架构的搜索空间,使得在GPU天数方面,在可能架构的集合中发现最佳候选者是易处理的和高效的。该方法消除了对于给定目标任务从头开始训练网络的要求,并且在搜索空间的初始化中引入了来自人类领域的偏差。

英文摘要:

In this paper, we propose a neural architecture search framework based on a similarity measure between some baseline tasks and a target task. We first define the notion of the task similarity based on the log-determinant of the Fisher Information matrix. Next, we compute the task similarity from each of the baseline tasks to the target task. By utilizing the relation between a target and a set of learned baseline tasks, the search space of architectures for the target task can be significantly reduced, making the discovery of the best candidates in the set of possible architectures tractable and efficient, in terms of GPU days. This method eliminates the requirement for training the networks from scratch for a given target task as well as introducing the bias in the initialization of the search space from the human domain.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑