Alan M. Polansky, Ph.D.
Most of Alan Polansky’s work has centered around two modern nonparametric methods: the bootstrap and smoothing methods. Nonparametric statistical methods attempt to find valid methods for statistical inference without making many assumptions about the underlying population. The bootstrap is a general methodology, developed by Bradley Efron in 1977, that replaces analytical calculations with computer-based simulations. Smoothing methods seek to replace parametric assumptions like linearity in regression with less restrictive assumptions such as the regression curve being differentiable.
Later, Alan developed an alternative approach to statistical problems that are usually solved using multiple comparison techniques. These methods involve a measure of confidence he developed in his 2007 book, Observed Confidence Levels: Theory and Application, published by CRC/Chapman and Hall.
His interest in asymptotic statistics culminated in the publication of his book, Introduction to Statistical Limit Theory, also published by CRC/Chapman and Hall.
Since that time, he has become interested in statistical analysis of data arising from networks and is currently considering Bayesian inference for stochastic processes on networks.
Alan joined NIU in 1995.
Akakpo, R. M., Xia, M., and Polansky, A. M. (2019). Frequentist inference in insurance ratemaking models adjusting for misrepresentation. ASTIN Bulletin - The Journal of the International Actuarial Association.
Polansky, A. M., and Ghosh, S. (2016). Using observed confidence levels to perform principal component analyses. Communications in Statistics Series A: Theory and Methods, 45, 3596-3611.
Ghosh, S., and Polansky, A. M. (2016). New bootstrap confidence intervals for means of positively skewed distributions. Communications in Statistics Series A: Theory and Methods, 45, 6915-6927.
Polansky, A. M., and Maple, A. (2014). Using Bayesian models to assess the capability of a manufacturing process in the presence of unobserved assignable causes. Quality Technology and Quantitative Management, 13, 139-164.
Ghosh, S., and Polansky, A. M. (2014). Smoothed and iterated bootstrap confidence regions for parameter vectors. Journal of Multivariate Analysis, 132, 171-182.
Polansky, A. M. (2011). Introduction to statistical limit theory. Boca Raton, FL: Chapman and Hall/CRC Press.
Polansky, A. M. (2007). Observed confidence levels: Theory and applications. Boca Raton, FL: Chapman and Hall/CRC Press.
Polansky, A. M. (2007). Detecting change-points in Markov chains. Computational Statistics and Data Analysis, 51, 6013-6026.
Polansky, A. M. (2003). Selecting the best treatment in designed experiments. Statistics in Medicine, 22, 3461-3471.
Polansky, A. M. (2000). A smooth nonparametric approach to multivariate process capability. Technometrics, 43, 199-211.
Polansky, A. M., and Schucany, W. R. (1997). Kernel smoothing to improve bootstrap confidence intervals. Journal of the Royal Statistical Society, Series B, 59, 821-838.
Honors Faculty Fellow (2022-2023)
As an Honors Faculty Fellow, Professor Polansky will teach a seminar on Data and Social Justice in fall 2022 in the University Honors Program. The Honors Faculty Fellowship program identifies faculty eager to teach innovative, exciting seminars of interest to highly-motivated students from across the university.