Publication | Closed Access
The Marginal Value of Adaptive Gradient Methods in Machine Learning
246
Citations
0
References
2017
Year
Artificial IntelligenceMarginal ValueEngineeringMachine LearningData ScienceAdaptive Optimization MethodsSparse Neural NetworkSupervised LearningComputational Learning TheoryMachine Learning ModelLarge Scale OptimizationComputer ScienceAdaptive AlgorithmStatistical Learning TheoryDeep LearningNeural Architecture SearchFunctional Data AnalysisEmpirical Generalization CapabilityAdaptive OptimizationLocal OptimizationStatistical Inference
Adaptive optimization methods, which perform local optimization with a metric constructed from the history of iterates, are becoming increasingly popular for training deep neural networks. Examples include AdaGrad, RMSProp, and Adam. We show that for simple overparameterized problems, adaptive methods often find drastically different solutions than gradient descent (GD) or stochastic gradient descent (SGD). We construct an illustrative binary classification problem where the data is linearly separable, GD and SGD achieve zero test error, and AdaGrad, Adam, and RMSProp attain test errors arbitrarily close to half. We additionally study the empirical generalization capability of adaptive methods on several state-of-the-art deep learning models. We observe that the solutions found by adaptive methods generalize worse (often significantly worse) than SGD, even when these solutions have better training performance. These results suggest that practitioners should reconsider the use of adaptive methods to train neural networks.