Paul Anderson, M.M. Meerschaert and Kai Zhang, Forecasting with prediction intervals for periodic autoregressive moving average models, Journal of Time Series Analysis, Vol. 34 (2013), No. 2, pp.187-193. Click here for the R codes and data.
Periodic AIC and model selection for periodic ARMA model. My main theorem proves that the periodic AIC has the following form.
The reduced model for periodic ARMA. The parsimony of periodic ARMA reduced models is obtained from two asymptotic distributions,
Led 34 recitation classes for undergraduate Business Statistics course; Instructed a total of 1020 students, with the average teaching evaluation of 1.5/5; (1=Excellent, 5=Poor.) Independently lectured Business Statistics course in summer of 2010; instructed 30 students, with the teaching evaluation of 1.1/5.
Wrote and tested R codes for model fitting and diagnostics for historical river flow data;
Forecasted the future 24-month river flows, with 95% prediction intervals;
Supported by NSF grants EAR-0823965, PI: Dr. Mark Meerschaert
Collected and analyzed survey data, based on 1500 questionnaires from national customers;
Provided text analysis by SPSS, graphic and table report by Excel.
11/2010, Department of Statistics and Probability, MSU
Awarded for the best performance of teaching assistants annually (2 out of 35)
See here for the departmental news.
05/2012, College of Natural Science, MSU
Awarded for the stimulus of full-devoted research in summer
SAS Certified Base Programmer for SAS 9, License BP028516v9, 11/2012.
Exam P, awarded by Society of Actuaries, 06/2010.