arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 1810.08178cs.LGcs.AIstat.ML
收藏

梯度一致性作为元学习的优化目标

Gradient Agreement as an Optimization Objective for Meta-Learning

University of Southern California(南加州大学);Clarifai(Clarifai公司)

Amir Erfan Eshratifar, David Eigen, Massoud Pedram

摘要:

本文提出了一种新的优化方法,以最大限度地提高概括的任务在元学习。元学习的目标是学习一个模型,使智能体在面对以前看不见的任务时能够快速适应。任务是从一个特定的分布中采样的,该分布被假设为对于可见和不可见的任务都是相似的。我们专注于一个家庭的元学习方法学习初始参数的基础模型,可以微调快速的一个新的任务,通过几个梯度步骤(MAML)。我们的方法是基于将模型的参数推向任务更一致的方向。如果任务的梯度与参数更新向量一致,则它们的内积将是一个大的正值。因此,给定一批要优化的任务,如果任务的梯度与该批中所有任务的梯度平均值之间的内积为正(负)值,则将正(负)权重与任务的损失函数相关联。因此,通过在任务的损失函数上引入一组权重来控制任务对参数更新的贡献程度。我们的方法可以很容易地集成到当前的元学习算法的神经网络。我们的实验表明,与MAML和Reptile相比,它产生的模型具有更好的泛化能力。

英文摘要:

This paper presents a novel optimization method for maximizing generalization over tasks in meta-learning. The goal of meta-learning is to learn a model for an agent adapting rapidly when presented with previously unseen tasks. Tasks are sampled from a specific distribution which is assumed to be similar for both seen and unseen tasks. We focus on a family of meta-learning methods learning initial parameters of a base model which can be fine-tuned quickly on a new task, by few gradient steps (MAML). Our approach is based on pushing the parameters of the model to a direction in which tasks have more agreement upon. If the gradients of a task agree with the parameters update vector, then their inner product will be a large positive value. As a result, given a batch of tasks to be optimized for, we associate a positive (negative) weight to the loss function of a task, if the inner product between its gradients and the average of the gradients of all tasks in the batch is a positive (negative) value. Therefore, the degree of the contribution of a task to the parameter updates is controlled by introducing a set of weights on the loss function of the tasks. Our method can be easily integrated with the current meta-learning algorithms for neural networks. Our experiments demonstrate that it yields models with better generalization compared to MAML and Reptile.

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。

↑