few-shot learning and meta learning

Humans can recognize new object classes from very few instances. However, most machine learning techniques require thousands of examples to achieve similar performance. The goal of few-shot learning is to classify new data having seen only a few training examples.

 N-way-K-shot classification. Here, we aim to discriminate between $\mathrm{N classes\; with K examples\; of\; each.}$

---

time series classification and forecasting 是可以用CNN做的，用的conv1D, 但是不知道和RNN相比效果怎么样？

References: 

1. Tutorial #2: few-shot learning and meta-learning I

2. “元学习”的理解

3. 模型无关的元学习：learn to learn

4. 关于Deep Learning未来发展的十大挑战（瓶颈）

5. How to Develop Convolutional Neural Network Models for Time Series Forecasting