This section defines what is probably the most common type of machine learning problem: that of supervised learning. I will formally define supervised learning later.
We use an example to introduce what is supervised learning and put the formal definition later. Let's say you want to predict house prices.
A while back, a student collected some data on house prices from the Oregon Institute in Portland. If you plot the data, it would look like this: the horizontal axis represents the area of the house in square feet, and the vertical axis represents the house price in thousands of dollars. Based on this set of data, if you have a friend who has a 750-square-foot house, and now he wants to sell the house, he wants to know how much the house can sell for.
So on this question, how will machine learning algorithms help you?
By applying a learning algorithm, we can draw a straight line in this set of data, or in other words, fit a straight line. Based on this line, we can infer that this house may be sold. Of course, this is not the only algorithm. There may be better ones. For example, instead of fitting the data with a straight line, it may be better to use a quadratic equation to fit the data. According to the curve of the quadratic equation, we can infer from this point that the house can be sold close. Later we will discuss how to choose a learning algorithm and how to decide whether to use a straight line or a quadratic equation to fit. One of the two options will sell your friend's house more reasonably. These are all good examples of learning algorithms. The above is an example of supervised learning.
It can be seen that supervised learning refers to the fact that we give the learning algorithm a data set. This data set consists of "correct answers". In the example of house prices, we are given a series of house data. We are given the correct price of each sample in the data set, that is, their actual selling price, and then use the learning algorithm to calculate more correct answers. For example, the price of your friend’s new house. In jargon, this is called a regression problem. We try to infer the result of a continuous value, the price of a house.
ordinary house