The principle and state-of-art applications of permutation tests

Contemporarily, permutation test is widely used in plenty of types of verification in state-of-art bigdata analysis. In this paper, the principle of permutation tests will be demonstrated with corresponding applications in various fields. Generally speaking, the permutation test can be regarded as an exact exam. On this basis, it can be also treated as a type of statistical significance level verifications where the value of sample distributed among the null hypothesis and alternative hypothesis are derived according to resampling and calculating of all possible combinations of the element. In other word, permutation test can be also defined as a procedure of resampling. For the sake of explicating the applications and meaning of the test clearly, this paper will first introduce the basic descriptions and formulae of the test and illustrate the applications subsequently. Specifically, the result of principle of permutation test is constructing null hypothesis and significance level, then equation will be presented to calculate p-value to compare with the primary data to determine one whether rejects the null hypothesis or not. Besides, permutation tests’ extensive functions on biological, medical, and economical areas will be discussed. Based on the analysis, these results offer a better understanding on principle and applications of permutation tests.