# Square Test PPT: Formula, Characteristics, Limitations

Chi-Square Test PPT: Formula, Characteristics, Limitations

Machine learning is booming day by day and time has changed tremendously. Chi- square test applications in machine learning is making this test the best topic of discussion as everyone wants to know how chi- square test makes a difference in machine learning. In machine learning there has been always a problem of feature selection as if you have multiple solutions then how you will choose the best feature available to build your model. Here comes the role of chi-square test as it helps in finding the best solution by examining the relationship between the elements. Now lets us discuss chi-square test in detail, and we will learn its various characteristics, formula and limitations.

### Formula for Chi-Square Test

To determine the difference between observed and expected data, a statistical procedure is used that is called chi-square test. And its formula is written below

Where

0 is Observed Value

E is Expected Value

C is Degrees of Freedom

Degrees of freedoms are the numbers of variables, observed values are gathered results and expected values are the frequencies expected.

### Characteristics of Chi-Square Test

Chi-square test is totally dependent on the frequencies and not on the mean and standard deviation like parameters.

It is a hypothetical test and cannot be used as an estimation test.

It is very useful test in research work as it can be used to solve complex problems.

### Who uses Chi- square test?

This test is widely used by the researchers who are studying on surveys because it works on category variables. Economic, political and marketing decisions can be taken with the help of chi-square test.

### Limitations of Chi-Square Test

This test is for experienced researchers and not for the starters as it is extremely sensitive.

It requires a lot more detailed analysis as it only determines whether two variables are related.

So it was all about chi-square test and how we can use it in machine learning.