吴恩达深度学习笔记1-Course1-Week1【深度学习概论】

2018.5.7

吴恩达深度学习视频教程网址

网易云课堂：https://mooc.study.163.com/smartSpec/detail/1001319001.htm

Coursera：https://www.coursera.org/learn/neural-networks-deep-learning

PS：网易云上不提供测验和作业，Cousera上有。

--------------------------------------------------------------------------------

深度学习概论：

--------------------------------------------------------------------------------

本篇主要关于深度学习的一些介绍和几个相关的术语

--------------------------------------------------------------------------------

Single/Multiple neural network:

深度学习的本质：Given data (input and output) and  fit a function that will predict output. 

修正线性单元：（Rectified Linear Unit --ReLU)一种人工神经网络中常用的激活函数（activation function）

卷积神经网络：Convolution Neural Network (CNN) used often for image application 

循环神经网络：Recurrent Neural Network (RNN) used for one-dimensional sequencedata such as translating English to Chinses or a temporal component such astext transcript.

结构化数据：Structured data refers to things that has a defined meaning such as price, age 

非结构化数据：Unstructured data refers to thing like pixel, raw audio, text.

深度学习三要素： large amount of data available with label, fast computation and neural network algorithm.

Two things have to be considered to get to the high level of performance:
1. Being able to train a big enough neural network
2. Huge amount of labeled data

训练神经网络是一个迭代过程：Idear--Code--Experiment

It could take a good amount of time to train a neural network, which affects your productivity. Faster computation helps to iterate and improve new algorithm.