随机梯度下降算法
利用
梯度下降
下降(航空)
简单(哲学)
计算机科学
机器学习
人工智能
梯度法
趋同(经济学)
坐标下降
深度学习
人工神经网络
下降方向
算法
哲学
认识论
计算机安全
作者
Marcin Andrychowicz,Misha Denil,Sergio Luis Suárez Gómez,Matthew D. Hoffman,David Pfau,Tom Schaul,Brendan Shillingford,Nando de Freitas
出处
期刊:Cornell University - arXiv
日期:2016-06-14
卷期号:29: 3981-3989
被引量:584
摘要
The move from hand-designed features to learned features in machine learning has been wildly successful. In spite of this, optimization algorithms are still designed by hand. In this paper we show how the design of an optimization algorithm can be cast as a learning problem, allowing the algorithm to learn to exploit structure in the problems of interest in an automatic way. Our learned algorithms, implemented by LSTMs, outperform generic, hand-designed competitors on the tasks for which they are trained, and also generalize well to new tasks with similar structure. We demonstrate this on a number of tasks, including simple convex problems, training neural networks, and styling images with neural art.
科研通智能强力驱动
Strongly Powered by AbleSci AI