supervised learning
Supervised learning: This means that we give the learning algorithm a dataset. This dataset consists of " correct answers ".
Regression: Used to predict a series of continuous-valued attributes
Example 1. House price prediction
With the area as the abscissa and the housing price as the ordinate, determine several coordinate points.
We regard these points as continuous (Continuous), so a regression line can be drawn, but it is not unique, it can be a regression line or a regression curve, etc.
Classification: predicting discrete output values
Example 2. Prediction of breast cancer
Take the characteristic of the size of the tumor as the abscissa, and whether it is a malignant tumor as the ordinate, and determine several points.
We treat these points as discrete.
The criteria for judging breast cancer are not only the size of the tumor, but may also include age, thickness of the tumor, whether the size is uniform, whether the shape is uniform, and so on. There can be an infinite number of features, and then algorithms are used to solve them.
Example 3:
Given data on the size of houses in the real estate market, try to predict their prices. Price is a continuous output as a function of size, so this is a regression problem.
We can turn this example into a classification problem and instead make our output the house "sold for more or less than asking price". Here, we classify houses based on price into two discrete classes.
Example 4:
(a) Regression - Given a picture of a person, we have to predict their age given the picture
(b) Classification - Given a patient has a tumor, we have to predict whether the tumor is malignant or benign.