吴恩达机器学习笔记一（Machine learning algorithms）

Machine learning algorithms

（纯粹促进自己学习整理）

1. Machine learning algorithms:

Main two types:

Supervised learning 监督学习

In supervised learning, the idea is that we’re going to teach the computer how to do something.

Unsupervised learning 无监督学习

In unsupervised learning, we’re going let the computer learn by itself.

other: Reinforcement learning(强化学习)、 recommender system(推荐系统).

2. The definition of supervised learning

(Regression problem & Classification problem)

Example 1: Housing price prediction

“right answers” given

also called regression problem(回归问题)

Example 2: Breast cancer(malignant, benign)

Discrete valued output (0 or 1) discrete 英[dɪˈskriːt] 离散的; 分离的; 互不相连的; 各别的

Classification problem

When the features becomes two:

Maybe for some learning problems, you want to use an infinite number of features, an infinite number of attributes, so that your learning algorithm has lots of attributes of features or cues with which to make those predictions. So how do you deal with an infinite number of features? for example: the algorithm called the Support Vector Machine(支持向量机算法)

3. The definition of unsupervised learning

(clustering algorithm & Cocktail party problem)

Example 1:

there is a data set, but they all has the same label or really no labels, can you find some structure in the data? Given the data set, an Unsupervised Learning algorithm might decide that the data lives in two different clusters.（not given the right answer）

clustering algorithm(聚类算法)

Example 2: cocktail parties(鸡尾酒算法)

Imagine there’s a party, room full of people, all sitting around, all talking at the same time. And there are all these overlapping voices, because everyone is talking at the same time, and it is almost hard to hear the person in front of you. So maybe at a cocktail party with two people, two people talking at the same time, And we’re going to put two microphones in the room, And because these microphones are at different distances from the speakers, each microphone records a different combination of these two speaker voices.

So we can do, is take these two microphone recordings and give them to an Unsupervised Learning algorithm, called the cocktail party algorithm. and tell the algorithm find structure in this data for you, separate the sound.

actually, the cocktail party problem algorithm need one code:

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

this course we use Octave, Octave is free open source software.
Video link： https://www.bilibili.com/video/BV164411S78V?from=search&seid=9940536896957991242
中文笔记：http://www.ai-start.com/ml2014/