ProxyConvexity: AUnifiedFramework fortheAnalysisofNeuralNetworks TrainedbyGradientDescent

Open in new window