Current location - Education and Training Encyclopedia - Graduation thesis - The paper is irrelevant. What do you mean by removing extreme values?
The paper is irrelevant. What do you mean by removing extreme values?
Need to convert data.

The most common method is to check whether there are still problems in data processing, such as text conversion, extreme value removal, sorting and so on. , and then check whether descriptive statistics conform to common sense and get the right one.

Although this does not affect the meaning and direction too much, remember to add 1 if it is less than 1, because forgetting to add 1 has made the opposite result.

Look at the result after a basic operation. If there is no expected result, you can increase or decrease the control variables, fine-tune the model, or combine various methods (such as grouping and cross multiplication, the result is better).

. In short, as long as the theoretical derivation is reasonable, you can always get the desired results, but if you really can't get out, don't get stuck in a dead end or add stars yourself. When changing variables, whether they are all irrelevant or individual irrelevant, we must first process, log, standardize and so on.