神經網絡學習率

在吳恩達作業，第一課第二週中，有這樣一個例子   Interpretation: Different learning rates give different costs and thus different predictions results. If the learning rate is too large (0.01), the cost may oscillate up and d

>>阅读原文<<