Good evening. The Airpots loss incident (found) happened earlier, and a paragraph was added.
Today I would like to summarize the classification and regression in machine learning.
The main purpose of classification is ** to predict the class (Yes, No) to which the data belongs **. If the number of classes to be predicted is two, it is called ** binary classification **. ex) --Determine whether or not to withdraw a subscription from the user's usage history. ――Whether the student's profile passes or fails --Email information, is it spam or spam?
If there are more than two classes, it is called a ** multiclass classification **. ex) --Student evaluation --Image recognition
The main purpose of regression is ** value prediction **. More deeply, the task of explaining ** explanatory variables (features) ** to ** objective variables **. ex) --Explanation of math scores with science scores. (Explanatory variable: science score, objective variable: math score) --Explanation of rent by property size and area. (Explanatory variables: size and area, objective variables: rent)
And if you want to embark on linear regression, Assuming that the objective variable is y and there are n explanatory variables, $ x_1 $, $ x_2 $, ..., $ x_n $,
y = a_0 + a_1x_1 +・ ・ ・+ a_nx_n
, $ A_0, a_1, ···, a_n $. In order to obtain these, there are ** maximum likelihood method ** and ** least squares method **. Also, when $ n = 1 $, it is called ** simple regression **, and the others are called ** multiple regression **.
Let me give you an example of the references below. Suppose you have information about a restaurant that your customer has visited in the past. Classification: Predict whether you like or don't like new restaurants Regression: Predict how many times you will visit a new restaurant in the future
Finally, I would like to mention two things in common. Both are "** supervised learning **". By the way, supervised learning is learning using the correct label attached to the data.
It's easy, but I've summarized the classification regression. Thank you very much.
Differences between classification and regression for beginners in machine learning
Recommended Posts