Journal of Advanced Statistics

Download PDF (401.4 KB) PP. 181 - 190 Pub. Date: December 1, 2016

DOI: 10.22606/jas.2016.14001

• Zhenmin Chen^*
  Department of Mathematics and Statistics, Florida International University, Miami, FL 33199, U.S.A
• Zihao Li
  Department of Mathematics and Statistics, Florida International University, Miami, FL 33199, U.S.A
• Min Zhou
  Department of Mathematics and Statistics, Florida International University, Miami, FL 33199, U.S.A

The purpose of this research is to propose a new method for detecting change points with an epidemic alternative (in the form of a step function). There are several parametric approaches and nonparametric approaches in the literature that can be used for detecting change-points in epidemic models. Yan [16] summarized some existing parametric approaches. The approaches summarized in Yao’s paper are based on the assumption of known population variances. The proposed test statistic in this research does not depend on the assumption of known population variances. This better fits the real world situation. Monte-Carlo simulation was used to find the critical values of the test. The power study was also based on Monte-Carlo simulation. The simulation result shows that the test statistic proposed in this research provides quite decent power compared with other existing statistical procedures, especially for the case that the step is large and duration of the epidemic is long. In 1997, a likelihood ratio test was proposed by Csorgo and Horvath [24]. Compared with the likelihood ratio test, the method proposed in this paper is easier to use by the statistics users.

Change-point, epidemic alternative, power comparison, Monte Carlo simulation, unknown variances.

[1] Page ES. Continuous inspection schemes. Biometrika. 1954; 41: 100-155.

[2] Page ES. A test for a change in a parameter occurring at an unknown point. Biometrika. 1955; 42: 523-527.

[3] Chernoff H, Zacks S. Estimating the current mean of a normal distribution which is subjected to changes over time. Annals of Mathematical Statistics. 1964; 35: 999-1018.

[4] Kander Z, Zacks S. Test procedures for possible changes in parameters of statistical distributions occurring at unknown time points. Annals of Mathematical Statistics. 1966; 37: 1196-1210.

[5] Sen A, Srivastava MS. Some one-sided tests for change in level. Technometrics. 1975; 17: 61-64.

[6] Worseley KJ. Confidence regions and tests for a change-point in a sequence of exponential family random variables. Biometrika. 1986; 73: 91-105.

[7] Chen J, Gupta AK. Likelihood procedure for testing change-points hypothesis for multivariate Gaussian model. Random Operators and Stochastic Equations. 1995; 3: 235-244.

[8] Chen J, Gupta AK. Testing and locating variance change-points with applications to stock prices. Journal of the American Statistical Association. 1997; 92: 739-747.

[9] Gombay E, Hovarth L. An application of the maximum likelihood test to the change-point problem. Stochastic processes and applications. 1994; 50: 161-171.

[10] Gombay E, Hovarth L. Approximations for the time of change and the power function in change-point models. Journal of Statistical Planning and Inference. 1996; 52: 43-66.

[11] Levin B, Kline J. The cusum test of homogeneity with an application in spontaneous abortion epidemiology. Statistics in Medicine. 1985; 4: 469-488.

[12] Commenges D, Seal J, Pinatal F. Inference about a change-point in experimental neurophysiology. Mathematical Biosciences. 1986; 80: 81-108.

[13] Broemeling LD, Tsurumi H. Econometrics and Structural Change. New York: Marcel Dekker; 1987.

[14] Siegmund D. Approximate tail probabilities for the maxima of some random fields. Annals of Probability. 1988; 16: 67-80.

[15] Siegmund D. Confidence sets in change-point problems. International Statistical Review. 1988; 56: 31-48.

[16] Yao Q. Tests for change-points with epidemic alternatives. Biometrika. 1993; 80: 179-191.

[17] Guan Z. Semiparametric tests for change-points with epidemic alternatives. Journal of Statistical Planning and Inference. 2007; 137: 1748-1764.

[18] Pettitt AN. A simple cumulative sum type statistic for the change-point problem with zero-one observations. Biometrika. 1980; 67: 79-84.

[19] Brown RT, Durbin J, Evans JM. Techniques for testing the constancy of regression relations over time. Journal of the Royal Statistical Society, Series B (Methodological). 1975; 37: 149-192.

[20] Hogan ML, Siegmund D. Large deviations for the maxima of some random fields. Advances in Applied Mathematics. 1986; 7: 2-22.

[21] Siegmund D. Boundary crossing probabilities and statistical applications. Annals of Statistics. 1986; 14: 361-404.

[22] Yao Q. Large deviations for boundary crossing probabilities of some random fields. Journal Mathematical Research & Exposition. 1989; 9: 181-192.

[23] Yao Q. Boundary crossing probabilities of some random fields related to likelihood ratio tests for epidemic alternatives. Journal of Applied Probabilities. 1993; 30: 52-65.

[24] Csorgo M, Horvath L. Limit Theorems in Change-Point Analysis. Chichester: John Wiley & Sons; 1997.