Faculty

Title: Professor of School of statistics,Doctoral Advisor
Position:Vice Director of Centered Applied Statistics
Email: mztian@ruc.edu.cn

Education

B.S.  (1991), National University of Defense Technology, China;
       (1993), Hunan Institute of Education, China;
M.A. (1998), University of Hunan, China;
Ph.D.(2001), University of  Nan Kai, China.

Experience

• 2008.07-2009. 01, SFB fellow, Humboldt Univerisity,Germany
• 2008.03-2008.06, Visiting Research Scholar of Baptist University of Hong Kong, Hong Kong.
• 2007.08-2007. 11, Visiting Research Scholar of Baptist University of Hong Kong, Hong Kong.
• 2005.07~2007.2, Postdoctoral fellow, the Chinese University of Hong Kong and the Baptist University of Hong Kong, Hong Kong.
• 2004.11~2005.02, Associate researcher, the Chinese University of Hong Kong, Hong Kong.
• 2004.01 ~07, Postdoctoral fellow, Department of Mathematics and Statistics, University of Calgary, Canada.
• 2002.01 ~2003.12, Postdoctoral fellow, Canadian Center for Advanced Studies of National Databases, University of Alberta, Canada.
• 2002.08~.2002.11, Assistant Researcher, Chinese University of Hong Kong, Hong Kong.
• 2001.06~2004.08, Postdoctoral fellow, Academy of Mathematics and System Sciences, Chinese Academy of Sciences, China.

Publications

Books

• Wu, X. Z. and Tian, M. Z. (2003), Diagnostics for Modern Regression Models. China Statistical Press. (In Chinese)

Selected papers

• Tian, M. Z., Tang, M. L., Ng, H. K. T. and Chan, P. S. (2008), Confidence interval estimators for risk ratio under inverse sampling. Statistics in Medicine. 27: 3301–3324. (SCI)
• Tian, M. Z., Tang, M. L., Ng , H. K. T. and Chan, P. S. (2008), A comparative study of confidence intervals for negative binomial proportion. Journal of Statistical Computation and Simulation. 1-10. (SCI)
• Tang, M. L., Tian, M. Z.and Chan, P. S. (2008), On the bootstrap quantile-treatment-effect test. Journal of Applied Statistics. Vol. 35 (1), 335-350. (SCI)
• Tang, M. L. and Tian, M. Z. (2008), Asymptotic interval estimation of risk difference under inverse sampling Computational Statistics and Data Analysis， Ref. No.:  CSDA-D-07-00933, (In press, SCI)
• Tang, M. L. and Tian, M. Z. (2008), Exact Interval Estimation for The Risk Difference under Inverse Sampling, Biometrics , (Under revision), (SCI)
• Tian, M. Z., Tang, M. L., and Chan, P. S. (2008), Semiparametric quantile modelling of hierarchical data. Acta Mathematica Sinica (In press, Ref. B7220) , (SCI)
• Wu X. and Tian, M. Z., (2008), A longitudinal study of the effects of family background factors on mathematics achievements using quantile regression. Acta Mathematicae Applicatae Sinica (English Series). 24(1), 85-98. (SCI)
  Zhong, Y. and Tian, M. Z. (2008) Bayesian analysis of change-point problems in rare events. Statistics & Decision. Vol. 3, 38-43. (In Chinese)
• Luo, Y. B., Tian, M. Z.and Wu, X. Z., (2008) Saddlepoint approximations to generalized chi-squared mixed distributions. Statistics & Information Forum, Vol. 1, 29-32. (In Chinese)
• Tian, M. Z., Chan, N. H. (2008), Saddlepoint Approximations to VaR, ( Statistica Sinica, Under revision) (SCI)
• Tian, M. Z., Wu, X. Li, Y. and Zhou, P. (2007), Longitudinal study of the external pressure effects on children’s mathematics and science achievements using nonparametric quantile regression. Chinese Journal of Applied Probability and Statistics. 24, 327-336.
• Tian, M. Z., Li, Y., and Zhou, P. (2008), Approximate and asymptotic confidence intervals for epidemiologic rate under inverse sampling. Journal of Systems Science and Complexity. 28, 513–523.
• Tian, M. Z., Wu, X. Li, Y. and Zhou, P. (2008), An analysis of mathematics and sciences achievements of American youth with nonparametric quantile regression. Journal of Data Science, 6, 449-465
• Tian, M. Z. and Chen, G. M. (2006), Quantile-hierarchical models. Science in China Series A: Mathematics, 36(10), 1103-1118. (In Chinese)
• Tian, M. Z. (2006), A quantile regression analysis of family background factor effects on mathematical achievements, Journal of Data Science, 4, 28-47.
• Tian, M. Z. and Chen, G. M. (2006), Hierarchical linear regression models for conditional quantiles. Science in China Series A: Mathematics, 49, 11-16. (SCI)
• Tian,M. Z. (2006), Two stages inferences for a semi-parametric regression model. Acta Mathematicate Applicate Sinica, 29， 601-608
• Tian, M. Z. (2005), Extreme distribution of the weighted sum of a class of $m$ dependent stochastic variable sequences. Math. Theory Appl. 25, 5-9
• Tian, M. Z. (2005), Estimation theory based on quasi-residuals in sliced inverse regression, Journal of Systems Science and Mathematical Sciences, 25, 348-355.
• Tian, M. Z. and Li, G. Y. (2004), Quasi-residuals method in sliced inverse regression, Statistics and Probability Letters, 66, 205-211. (SCI)
• Tian, M. Z. and He, C. Z. (2003), A generalized variance-ratio test for a Heteroskedastic regression, Mathematics in Economics, 20, 52-61.
• Tian, M. Z. and Wu, X. Z. (2002), On an extended quasi-likelihood estimation and a diagnostic test for heteroscedasticity in the generalized linear models. Mathematical Theory and Applications, 22, 5-14.
• Tian, M. Z. and Wu, X. Z. (2001), A quasi-residuals method, Advances in mathematics, 30, 182-184.
• Tian, M. Z. and He, C. Z. (2001), Quasi-residual diagnostic theory and methodology for heteroscedastic model, Natur. Sci. J. Xiangtan Univ. 23, 1-8.
• Tian, M. Z. (2000), The limiting property of error variance in a semi-parametric errors-in-variances model, Journal of Hunan University, 27, 4-9.

Papers under revision

• Tian, M. and Haerdle, W. (2008), Locally varying bandwidth selection for conditional quantile regression
• Fan, J. Y. and Tian, M. Z. (2008), A new index of goodness-of-fit tests for hierarchical linear models, Statistics & Decision.
• Tian, M. Z., Li, Y., and Zhou, P. (2007), Approximate and asymptotic confidence intervals for epidemiologic rate under inverse sampling. Communications in Statistics -- Theory and Methods. (Reference: A06-395 )
• Tian, M. Z., Tang, M. L., and Chan, P. S. (2006), Saddlepoint approximations to conditional probability integral in meta analysis.
• Tian, M. Z. (2007), A limit distribution for the generalized Erdös-Kac Statistic.
• Tian, M. Z. and G, Chen (2007), A naive lack-of-fit test for semi-parametric regression models. Journal of Nonparametric Statistics. (Reference: S-2006-144).
• Tian, M. Z. and G, Chen (2007), A limit distribution for the maximum of weighted sums of m-dependent random variables.
• Tian, M. Z. (2007), A quantile regression analysis of family background factor effects on Mathematics achievements.
• Tian, M. Z. (2007), Does external pressure matter? A quantile regression technique.

Teaching

Postgraduate course

(Renmin University of China)

• Statistical Models (3 hours)
• Quantile Regression (2 hours)
• Hierarchical Regression Models (2 hours)
• Modern Statistical Theory and Methods (2 hours)
• Saddlepoint Approximations (2 hours)
• Statistical Estimation of Epidemiological Risk (2 hours) Statistical Analysis with Complex Data (2 hours)
• Frontiers in Statistics (2 hours)
• Computer Intensive Methods (2 hours)
• Quantitative Risk Management (3 hours)

Undergraduate course

(Renmin University of China)

• Statistical Computation (2 hours)
• Regression Analysis (3 hours)
• Statistical Diagnostics (3 hours)

Faculty Positions

Assistant, China Insurance University (1994)
Associate professor, Renmin University of China (2004)
Full professor, Renmin University of China (2008).

Honors and Awards

2008 Recipient of the second prize for excellent scientific research fruits of the National Chinese Statistical Association.
2006 Excellent person of the Ministry of Education of the People’s Republic of China in the new century.
2006 Recipient of the second prize for excellent scientific research fruits of the National Chinese Statistical Association.
1991-1992 Outstanding Teacher

1. Locally Varying Bandwidth Selection for Conditional Quantile Regression, C.A.S.E., Humboldt University, Germany. July 10, 2008
2. Semiparametric Quantile Regression Models For Hierarchical Data, International Statistics Forum ，Renmin University of China, Beijing, China, June 10-11, 2006
3. Confidence interval for epidemiologic rate based on Saddle-point approximations approach under inverse sampling. Hua Zhong Normal University, Wuhan, China. April 17, 2006.
4. On Hierarchical Nonparametric Quantile Regression. Research seminar, the Chinese University of Hong Kong, March 28, 2006.
5. Longitudinal study of the external pressure effects on children's mathematics. Presented at the 2004 American Educational Research Association Meeting in San Diego, April 12-16, 2004.
6. Quantile models: theory and practical guideline for empirical research in education. Research seminar, University of Alberta, May 1 - July 31, 2003.
7. A generalized variance-ratio test for a heteroskedastic regression.The Beijing--Tianjin Conference of Chinese Probability and Statistics, Renmin University of China, China, May 20, 2002.
8. Quasi-residuals method in sliced inverse regression. The7-th Conference of Chinese Probability and Statistics in NortheastNormalUniversity, Sept, China, 19-24, 2002.
9. Statistical diagnostics for heteroscedasticity and outliers.
   Research seminar, Academy of Mathematics and System Sciences, ChineseAcademy of Sciences, China, July 25, 2001.
10. Quasi-residual diagnostic theory and methodology for heterosckedastic model. The 6-th Conference of Chinese Probability and Statistics, Beijing, China, Sept. 19-24, 2001.

Professional Membership

•  The Chinese Society of Probability and Statistics
• American Educational Research Association
• Referee for more than 20 Journals:
  Communications in Statistics--Theory and Methods;
  Science in China:Acta Mathematicate Applicate Sinica;
  Systems Science and Complexity, etc.

Computer Skills

S-PLUS,MATLAB, MLwiN, HLM, SPSS, R, MATHEMATICA.
Microsoft Office, Latex.

Current Research Interests

Adaptive smoothing; Bayesian statistical infernce; Computer intensive methods; Data analysis; Extremes and heavy tails; Functional data analysis; Financial econometrics and risk management; Hierarchical models; Hierarchical- quantile regression modeling; High dimension reduction; Inverse problems; Large sample theory; Model selections; Nonparametric and semiparametric modeling; Order statistics; Quantile regression; Quantitative finance; Robust statistics; Saddlepoint approximations with applications; Statistical diagnostics; Statistical modethods in epidemiological risk; Stochastic simulations; Time series modeling; Volatility modelling.

Current cooperative projects abroad

The Project 1 (with Prof. Dr. Wolfgang Haerdle and Vladimir Spokoiny, Humboldt university ):
Locally adaptive methods for the modern quantile regression.

The Project 2 (with Prof. Dr. Peter Hall, University of Melbourne):
Background: The projects are all in nonparametric statistics, and include topics in functional data analysis, inverse problems and classification, variable selection and hypothesis testing for very high dimensional data. Combinations of these areas, for example classification methods for functional data, or variable selection in the presence of errors in variables, are also included.
Objectives:
(a) Development of new statistical methodology in the fields described above;
(b) Application of the methodology to real data;
(c) Development of theoretical tools appropriate to describing the performance of the methodology in (a), and development of appropriate optimality theory;
(d) Exploration of numerical performance of the methodology, through simulation studies.

Current grants

• The National Natural Science Foundation of China (No. 10871201)
• Education Ministry Key Project (No. 108120)
• National Philosophy and Social Science Foundation grant (No. 07BTJ002)
• New Century Excellent Talent Program

Dr. Tian is particularly interested in developing appealing statistical approaches and their applications. He also takes delight in sophisticated statistical models. In addition, he has been made interdisciplinary collaborations with others for several years.