Adam: A Method for Stochastic Optimization
Diederik P. Kingma, Jimmy Lei Ba · arXiv (Cornell University) · 2014
We introduce Adam, an algorithm for first-order gradient-based optimization of stochastic objective functions, based on adaptive estimates of lower-order moments. The method is straightforward to implement, is computationally efficient, has little memory requirements, is invariant to diagonal rescaling of the gradients, and is well suited for problems that are large in terms of data and/or parameters. The method is also appropriate for non-stationary objectives and problems with very noisy and/or sparse gradients. The hyper-parameters have intuitive interpretations and typically require little tuning. Some connections to related algorithms, on which Adam was inspired, are discussed. We also analyze the theoretical convergence properties of the algorithm and provide a regret bound on the convergence rate that is comparable to the best known results under the online convex optimization framework. Empirical results demonstrate that Adam works well in practice and compares favorably to other stochastic optimization methods. Finally, we discuss AdaMax, a variant of Adam based on the infinity norm.
Cited by 79,368 works. doi:10.48550/arxiv.1412.6980 Open-access PDF
Authors
References
- Reducing the Dimensionality of Data with Neural Networks
- Improving neural networks by preventing co-adaptation of feature detectors
- Adaptive Subgradient Methods for Online Learning and Stochastic Optimization
- Deep Neural Networks for Acoustic Modeling in Speech Recognition: The Shared Views of Four Research Groups
- On the importance of initialization and momentum in deep learning
- Speech recognition with deep recurrent neural networks
- Generating Sequences With Recurrent Neural Networks
- ADADELTA: An Adaptive Learning Rate Method
- Acceleration of Stochastic Approximation by Averaging
- Learning Word Vectors for Sentiment Analysis