How to iron out rough landscapes and get optimal performances: Replicated Gradient Descent and its application to tensor PCA

Open in new window