Curriculum Vitae

Zhiqiang Cao

Contact Information

Email: caozhiqiang@sztu.edu.cn; Personal website: https://statczq.rbind.io/.

Research Interests

Survival analysis, Time series analysis, Statistical learning, Measurement error models.

Education Background

  • 2015.09 ~ 2019.05, PhD in Statistics, Department of Mathematics, The Hong Kong University of Science and Technology. Supervisor: Prof. Man-Yu Wong.
  • 2012.09 ~ 2015.07, Master of Probability Theory and Mathematical Statistics, School of Mathematical Science, Beijing Normal University. Supervisor: Dr. Hui Li.
  • 2008.09 ~ 2012.07, Bachelor of Statistics, School of Mathematics and Statistics, Chongqing University.

Honors & Awards

  • The Third Prize of Global Development Challenge (AI theme), Huawei Technologies Co. Ltd, 2019
  • Epsilon Fund Award, HKUST, 2019
  • The George K Lee Award, HKUST, 2019
  • Research Travel Grant (three times), HKUST, 2017-2018.
  • Best Oral Presentation Award: “Tests for Zero Inflation in Count Data with Measurement Error in Covariates”. 19th International Conference on Statistics and Analysis, Singapore, 2017.
  • Din-Yu Hsieh Teaching award, HKUST, 2016.
  • Excellent teaching assistant award, Beijing Normal University, 2015.
  • The Second-class Scholarship, Beijing Normal University. 2014.
  • Third Prize in National Undergraduate Statistical Modeling Contest, Statistics Education Society of China. 2011.
  • Third Prize in the Chinese Mathematics Competitions (Division of Chongqing, Math class), Chinese Mathematical Society. 2010.
  • National Encouragement Scholarship, Ministry of Education of P.R.C. 2010.
  • National Scholarship, Ministry of Education of P.R.C. 2009.
  • Various scholarships and awards from Chongqing University. 2008-2012.

Publications and Presentations

In English

Papers:

  1. Zhiqiang Cao, Man-Yu Wong. (2021). Approximate maximum likelihood estimation for logistic model with covariate measurement error. Biometrical Journal, 63(1), 27-45. URL:https://onlinelibrary.wiley.com/doi/full/10.1002/bimj.202000024

  2. Hui Li, Zhiqiang Cao, Guosheng Yin. (2018). Varying-association copula models for multivariate survival data. Canadian Journal of Statistics, 46(4), 556-576. URL:https://onlinelibrary.wiley.com/doi/10.1002/cjs.11474

  3. Zhiqiang Cao, L.L.Hui, M.Y.Wong. (2018). New approaches to obtaining individual peak height velocity and age at peak height velocity from the SITAR model. Computer Methods and Programs in Biomedicine, 163,79-85. URL:https://www.sciencedirect.com/science/article/pii/S0169260717313895

  4. M.Y.Wong, Yingsi Yang, Zhiqiang Cao, Vivian Y.W.Guo, Cindy L.K.Lam, Carlos K.H.Wong. (2018). Effects of health-related quality of life on health service utilization in patients with colorectal neoplasms. European Journal of Cancer Care, 27(6),e12926. URL:https://onlinelibrary.wiley.com/doi/10.1111/ecc.12926

  5. Zhiqiang Cao, Hui Li, M.Y.Wong. (2018). Functional-coefficient autoregressive and linear regression mixed model for nonlinear time series. Proceedings of 33th International Workshop on Statistical Modelling, 2,37-41. PDF

  6. Zhiqiang Cao, Man-Yu Wong. (2020). Moment estimation method of parameters in additive measurement error model, Computer Methods and Programs in Biomedicine, in press

  7. Zhiqiang Cao, Man-Yu Wong. (2020). Approximate profile likelihood estimation for Cox model with covariate measurement error, revision.

  8. Zhiqiang Cao, Man-Yu Wong. (2020). Reduced correlation coefficient estimation, under review.

  9. Zhiqiang Cao, Hui Li. (2021) Varying-association copula models for marginal additive hazard distribution, under review.

  10. Zhiqiang Cao, Hui Li, Guosheng Yin. (2021). Linear Regression and Varying-Coefficient Autoregressive Mixed Model, manuscript.

In Chinese

Papers:

  1. Zhiqiang Cao, Hui Li, Xingwei Tong. (2015). A long memory regression model of IF1407 contract and investment strategy. Journal of Beijing Normal University, Natural Science, 51(4):348-353.
  2. Zhiqiang Cao, Hui Li. (2015). Analysis of gold price fluctuations based on intervention model. Journal of Mathematics in Practice and Theory, 45(15):146-153.
  3. Zhiqiang Cao, Yang Wang, Wei Li.(2015). Aalen model and its application in medical research. Chinese Journal of Health Statistics, 32(2):335-337.
  4. Zhiqiang Cao, Yang Wang, Wei Li.(2015). Comparison of semi-parametric and nonparametric additive hazards models: with Tamiflu and traditional Chinese medicine to treat H1N1 flu effect for example. Chinese Journal of Health Statistics, 32(1):22-25.
  5. Jing Lu, Qin Liang, Zhiqiang Cao.(2012). Efficiency evolution of property insurance industry in China based on three-stage data envelopment: Analysis of Unbalance Panel Data from 2004 to 2009. Insurance Studies, (5):23-35.
  6. Zhiqiang Cao, Hong Liu, Qin Liang.(2012). An empirical study of the life insurance market efficiency based on CCA-three-stage DEA model. Application of statistical models: The 2011 national undergraduate statistical modeling contest winners selected papers. 437-472. Beijing: China Statistics Press.

Presentations

  1. Functional-Coefficient Autoregressive and Linear Regression Mixed Model for Nonlinear Time Series. The 33rd International Workshop on Statistical Modelling 2018, Bristol, UK, 2018.07
  2. Partial Varying-coefficient Regression and Autoregressive Model for Nonlinear Time Series. 38th International Symposium on Forecasting, Boulder, Colorado, USA, 2018.06
  3. Functional-Coefficient Autoregressive and Linear Regression Mixed Model. PhD seminar of Department of Mathematics, HKUST, Hong Kong, 2018.04
  4. Tests for Zero Inflation in Count Data with Measurement Error in Covariates. The ICSA 2017: 19th International Conference on Statistics and Analysis, Singapore, 2017.09
  5. Varying-Association Copula Models for Multivariate Survival Data. PhD seminar of Department of Mathematics, HKUST, Hong Kong, 2017.04
  6. The introduction of AFT model. Medical Research Statistical Center, Fu Wai Hospital, Beijing, 2013.11

Project Experience

  1. Exploratory methods in development work using EPIC-InterAct data, HKUST. 2017.07-2019.03

    • Involved in developing statistical methods to deal with covariate measured with error under logistic regression model and Cox regression model based on EPICInterAct data, especially for covariates have relative large measurement error, for example, FFQ (Food Frequency Questionnaire) data in nutrition.
    • Performance: We have derived moment method and approximation maximumlikelihood method to deal with covariates measured with error under logistic regression and Cox regression. Simulation results are as expected. Compared to naive and regression calibration methods, real data analysis from our methods are more reasonable.

  2. Statistical modelling of growth and development in assessing the determinants of pubertal timing and growth, Cooperation project between HKUST and School of Public Health, the University of Hong Kong. 2016.01-2017.11

    • Involved in building statistical models to analyze determinants of pubertal timing and growth based on “Children in 1997” data. Especially in developing methods to compute individual age at peak velocity (apv), peak velocity (pv) and height at peak velocity (ypv), since pv and apv are useful and objective markers for respectively the velocity and timing of puberty for both boys and girls.
    • Performance: Based on the SITAR model, we proposed the numerical method and the quadratic function method to calculate apv, pv and ypv. Our paper has been published on Journal–Computer Methods and Programs in Biomedicine. In addition, corresponding R package– iapvbs has also been created by us.

  3. Analysis of an ongoing large study on diabetes, School of Public Health, The University of Hong Kong. 2016.10-2017.04

    • Based on more than 600 thousands sample data of Hong Kong, taking part in building survival model to analyze relationship between mortality of Diabetes and related factors.
    • Performance: According to cross-validation, our proposed model performes better than the Joint Asia Diabetes Evaluation (JADE) model conducted by the Hong Kong Diabetes Registry, and the C-statistics of our model is larger. What’s more, we found some new significant factors, which were not included in the JADE model.

  4. Regression model based on survival data study, National Natural Science Foun- dation, Beijing Normal University. 2013.09-2014.10

    • Involved in programming and doing numerical simulation.

  5. Structure and efficiency of Chinese insurance market study,The Second Chongqing University College Students’ Scientific Research Training Program, Chongqing University. 2010.07-2011.07

    • Co-investigator, mainly responsible for building Three-stage DEA model and analyzing data, involved in writing the research paper.
    • Performance: The research paper was published, and our project got the first prize among all 550 final reports (top 1%).

Work Experience

Full-time Job

  1. Assistant Professor, College of Bigdata and Internet, Shenzhen Technology University, Shenzhen, 2020.10-present

    • Teaching and research.

  2. Research Engineer, Algorithm and Technology Development Department, Huawei Technologies Co. Ltd, Dongguan, 2019.06-2020.09

    • Taking part in AutoML group, mainly mining time series pattern of KPI data from AIOps.

Internship

  1. R&D Centre, Nanjing mint co., LTD, Nanjing, 2014.11-2015.04 (nearly Full-time)

    • Taking part in building models and writing programs to analyze image of value added tax invoice produced.
    • Performance: Based on the sample image of real invoice produced by the machine, the program found some characteristics of the value added tax invoice. For example, first fixed position of character “No”, then located interested number area and gray processed can save store space. These foundings were helpful in further developing programs for feature extraction and visual inspection.

  2. Medical Research Statistical Center, Fu Wai Hospital, Beijing. 2013.07-2013.11 (nearly Full-time)

    • Using additive hazard model to analyze the treatment effect of H1N1 Flu between traditional Chinese medicine and Tamiflu; taking part in PURE (Prescriptive Urban and Rural Epidemiology Study) project, analyzing relative data using descriptive statistics.
    • Performance: Based on the real data of H1N1 Flu in China, we applied the Aalen model to study relationship between mortality of H1N1 Flu and related factors from another perspective, and published relative results in the Chinese journal of health statistics. What’s more, we introduced how to use R software to analyze the Aalen model for interested medical statistical researchers.

Teaching Assistant

  1. HKUST, Hong Kong. 2016.2-2018.12
    • Survival analysis, Statistical inference, Regression analysis.
  2. Beijing Normal University, Beijing. 2013.09-2014.06
    • Time series analysis, Probability theory and mathematical statistics.

Vounteer

  1. The 7th R Conference of China, Beijing. 2014.04-2014.05
    • Participating in production meeting manual and analyzing data about participants.
    • Meeting on-site service.

R packages

  1. iapvbs:Individual Age at Peak Velocity Based on SITAR.
  2. gmusim:Growth Modelling using a Shape Invariant Model with Random Effects.

Computer Skills

  1. Programing: R(excellent), Matlab(good), C(good), SAS(basic), Python(basic).
  2. Others: Office software, Latex, Markdown.

Hobbies

  1. Listening to music
  2. Travelling
  3. Hiking
  4. Swimming.