Bio Research Teaching Vitae

Research Education Publications Conferences Teaching Awards Service & Experience Skills

Research Interest

My research interests and plans span the area of
Nonparametric Estimation and Multiple Comparisons,
as well as the problems in Predictions in different settings.

Education

Present

Post-doctoral Researcher University of Haifa. Haifa, Israel.
Research Topic: "Nonparametric Estimation."
Mentor: Dr. Alexander Goldenshluger

2019

Ph.D. Statistics University of South Carolina. Columbia, SC.
Dissertation Title: "Investigations on Multiple Interval Estimators."
Advisor: Dr. Edsel A. Peña

2013

M.S. Statistics Rutgers, the State University of New Jersey. New Brunswick, NJ

Publications & Revised/Resubmitted

2021+

6. Kim, T., Lieberman, B., Luta, G., and Peña, E., On Optimal Correlation-Based Prediction. The American Statistician. [Revised and Resubmitted]

2021

5. Kim, T., Lieberman, B., Luta, G., and Peña, E., Prediction Regions for Poisson and Over-Dispersed Poisson Regression Models with Applications to Forecasting Number of Deaths during the COVID-19 Pandemic. Open Statistics. [Accepted for Publication]

4. Alexander Goldenshluger and Taeho Kim, Density Deconvolution with Non-Standard Error Distributions: Rates of Convergence and Adaptive Estimatio, Electronic Journal of Statistics. Volume 15, Number 1 (2021), 3394-3427.

3. Kim, T., Lieberman, B., Luta, G., and Peña, E., Prediction Intervals for Poisson-Based Regression Model, Wiley Interdisciplinary Reviews, Computational Statistics. e1568.

2019

2. Edsel Peña and Taeho Kim, Median Confidence Regions in a Nonparametric Model, Electronic Journal of Statistics. Volume 13, Number 2 (2019), 2348-2390.

1. Taeho Kim and Edsel Peña, Improved Multiple Confidence Interval via Thresholding Informed by Prior Information, arXiv preprint. arXiv:1912.03756.

Manuscripts in Progress

3. Alexander Goldenshluger and Taeho Kim, Nonparametric System Identifications: This study considers a linear time-invariant system for continuous time. Based on a nonparametric approach with the Wiener process, the point-wise identification performances of impulse and frequency response functions are evaluated by the minimax criterion with their rates of convergence.

2. Kim, T., Lieberman, B., Luta, G., and Peña, E., Predictions based on Concordance Correlation Coefficients: This study investigates the sample-level inference of the prediction based on concordance correlation coefficients. The asymptotic distribution of the estimated predictor is derived for both single- and multiple-covariate. In addition, its performance is compared to that of the least-squares prediction.

1. Taeho Kim and Edsel Peña, General Thresdholding Approach in Multiple Interval Estimations: The thresholding approach in multiple interval estimation is initially built up for normal-normal model. This study attempts to lift the model restriction, incorporating other conjugate families, such as exponential-inverse gamma model and etc.