This blog post is a course note for Prof. Andrew Ng, and interested friends can find it in Coursera or NetEase open class.
Supervised learning
is an example of supervised learning regression analysis. The chart is designed to predict house prices.
For this scenario, the training set has a tag attribute. For example, each point in the diagram corresponds to the price of the house, and supervised learning is to look for patterns in these known training centers to predict house prices.
Unsupervised learning
There are several scenes of unsupervised learning.
Compared with supervised learning, unsupervised learning is the element of training set without tag attribute, and unsupervised learning should excavate valuable classification from these training centers. For example, the organization of computer clusters, through a number of properties of each machine, training which machines to coordinate work with each other, and put them into a class, the result of classification is trained to know after. For example, given a large number of images, to find a class of pictures and all other pictures of different processes, but also unsupervised learning.
As for classification and regression problems, the classification problem can be understood as limited training results, such as whether it is a disease, illness, or health. The regression problem is that the training results in a certain range, such as the above-mentioned price estimate problem, the price is within a range.
machinelearning----Lesson 1