When creating a model, the value you want to predict is called a target variable. It is also called a dependent variable or a reaction variable, but most intuitively you can think of it as the y value at y = f(x). In other words, it is a variable that the model targets for analysis. Target variables can be various values such as bank deposit results, movie review ratings, and graduate acceptance probability.
The categorical classification model and the continuous value classification model have target variables. It is important in these models to have a clear purpose, because it is essential to have a data set that is composed of independent input variables that are used to calculate prediction, and a target variable which is the prediction itself. For example, suppose you have quit smoking success rate as a target variable. You can predict the target variable of the model by calculating variables such as age, occupation, height, weight, and area of residence as explanatory variables.
Here is an example of predicting the probability of passing graduate school to help you understand better.
[There are a total of eight variables in the data set: GRE score, TOEFL score, university ranking, statement of purpose, letter of recommendation, undergraduate GPA, research experience, and admission probability.]
[Among the eight variables, the value the user wants to predict is the probability of admission, so the probability of admission is selected as the target variable.]
✓ After creating the model, you can fill in the predictor variables (X values) with the values you want to put and check the predicted value of the target variable.