吴恩达 ML 公开课笔记(1) - Supervised and Unsupervised Learning

This is my notes for the open course [Machine Learning](https://www.AndrewNg’s ML.org/learn/machine-learning/) from AndrewNg’s ML.

Supervised learning

0. Model

• given a set of data assigned with special features(experience)
• build a model through learning algorithm(task)
• predict the features through given data using the model built(performance)

1. Regression problem

• predict through consecutive data

2. Classification problem

• pretict between discrete data sets
• learning from multiple(even infinite) featurea as parameters

Unsupervised learning

0. Model

• given a set of data(with no features)
• build a model through learning algorithm like cluster, etc.
• classify though the model built, etc. (which means they might have some same features)

1. Cocktail party problem

• learning from a combination of different audio sources(in different ways, at least 2 ways to identify)
• separate different sources from the combiniton