Coursera-DeepLearning-Part3-Week2-take away

学习目标

• Understand what multi-task learning and transfer learning are
• Recognize bias, variance and data-mismatch by looking at the performances of your algorithm on train/dev/test sets

---

一. Error Analysis

1. Carry out error analysis

---

2. Cleaning up incorrectly labeled data

---

3. Building your first system quickly, then iterate

---

二. Mismatched training and dev/test set

1. Training and testing on different distributions

---

2. Bias and Variance with mismatched data distributions

---

3. Addressing data mismatch

---

三. Learning from multiple tasks

1. Transfer learning

补充材料：

---

2.Multi-task learning

---

四. End-to-end deep learning

1. What is end=to=end deep learning?

补充材料：

---

2. Whether to use end-to-end deep learning

---

Quiz:

只要人工合成的雾在人眼看来是真实的，您就可以确信，人工合成的数据能够准确地捕捉到真实雾图像的分布，因为人类的视觉对于您正在解决的问题是非常准确的