无人驾驶DeepReforcementLearning.pdf PhilosophicalMotivationforReinforcementLearningTakeawayfromSupervisedLearning:Neuralnetworksaregreatatmemorizationandnot(yet)greatatreasoning.HopeforR
learningRatesandHowItImprovesPerformanceinDeepLearning.pdf UnderstandingLearningRatesandHowItImprovesPerformanceinDeepLearningThispostisanattempttodocumentmyunderstandingonthefollowingtopic:Whatisthelearningra