机器学习（Machine Learning）- 吴恩达（Andrew Ng） 学习笔记（一）

Overview 总览

Machine learning

1. Grew out of work in AI

2. New capability for computers

Examples:

1. Database mining

   Large datasets from growth of automation/web.

   E.g., Web click data, medical records, biology, engineering

2. Applications can't program by hand.

   E.g., Autonomous helicopter, handwriting recognition, most of Natural Language Processing (NLP), Computer Vision.

3. Self-customizing programs

   E.g., Amazon, Netflix product recommendations.

4. Understanding human learning (brain, real AI).

What is machine learning

Machine Learning definition

1. Atrhur Samuel (1959). Machine Learning: Field of study that gives computers the ability to learn without being explicitly programmed.

2. Tom Mitchell (1998). Well-posed Learning Problem: A computer program is said to learn from experience E with respect to some task T and some performance measure P, if its performance on T as measured by P improves with experience E.

   一个程序被认为能从经验E中学习，解决任务T，达到性能度量值P，当且仅当有了经验E后，经过P评判，程序在处理T时的性能有所提升。

Machine learning algorithms

1. Supervised learning 监督学习

2. Unsupervised learning 无监督学习

Others

Reinforcement learning, recommender systems. 强化学习，推荐系统

Also talk about

Practical advice for applying learning algorithms. 我们的课程也会讨论对实际应用学习算法的建议

Supervised learning

Regression problem: 回归问题

1. Housing price prediction. 房价预测

2. You have a large inventory of identical items. You want to predict how many of these items will sell over the next 3 months. 商品销量预测

Classification problem: 分类问题

• Breast cancer (malignant, benign). 乳腺癌良性/恶性的区分
• You'd like software to examine individual customer accounts, and for each account decide if it has been hacked/compromised. 一些账户是否被黑过

Unsupervised learning

Some examples:

1. Organize computing clusters 组织计算集群（找出哪些适合协同工作）

2. Social network analysis 社交网络中人们关系的分析

3. Market segmentation 市场受众群体分析

4. Astronomical data analysis 航天数据分析

Cocktail party problem algorithm

（不懂，留坑）

[W, s, v] = svd((repmat(sum(x.\*x, 1), size(x, 1), 1).\*x)\*x')

Test:

Of the following examples, which would you address using an unsupervised learning algorithm?

1. Given email labeled as spam / not spam, learn a spam filter. × 根据对垃圾邮件的标注实现一个垃圾邮件过滤器

2. Given a set of news articles found on the web, group them into set of articles about the same story. √ 将文章按照类型进行分类

3. Given a database of customer data, automatically discover market segments and group customers into different market segments. √ 市场受众群体分析

4. Given a dataset of patients diagnosed as either having diabetes or not, learn to classify new patients as having diabetes or not. × 根据之前的数据集诊断病人是否有糖尿病