A Comprehensive Overview and Survey of Recent Advances in Meta-Learning

This article reviews meta-learning which seeks rapid and accurate model adaptation to unseen tasks with applications in image classification, natural language processing and robotics. Unlike deep learning, meta-learning uses few-shot datasets and concerns further improving model generalization to unseen tasks. Deep learning focuses upon in-sample prediction and meta-learning concerns model adaptation and out-of-sample prediction. Meta-learning may serve as an additional generalization block complementary for original deep learning model. Meta-learning can continually perform self-improvement to achieve highly autonomous AI. We summarize meta-learning models in the following categories: black-box adaptation, similarity based method and meta-learner procedure. Recent applications concentrate upon combination of meta-learning with Bayesian deep learning and reinforcement learning to provide feasible integrated problem solutions. We present performance comparison of recent meta-learning methods and discuss future research direction.