Provable Benefits of Unsupervised Pre-training and Transfer Learning via Single-Index Models
This provides theoretical insights for machine learning practitioners dealing with data scarcity, though it is incremental as it builds on existing pre-training methods.
The paper tackles the problem of high-dimensional supervised learning with limited labeled data by analyzing unsupervised pre-training and transfer learning, showing they reduce sample complexity by polynomial factors and sometimes grant exponential improvement over random initialization.
Unsupervised pre-training and transfer learning are commonly used techniques to initialize training algorithms for neural networks, particularly in settings with limited labeled data. In this paper, we study the effects of unsupervised pre-training and transfer learning on the sample complexity of high-dimensional supervised learning. Specifically, we consider the problem of training a single-layer neural network via online stochastic gradient descent. We establish that pre-training and transfer learning (under concept shift) reduce sample complexity by polynomial factors (in the dimension) under very general assumptions. We also uncover some surprising settings where pre-training grants exponential improvement over random initialization in terms of sample complexity.