Ming-Hui Chen

Overview

Ming-Hui Chen is a statistician (University of Connecticut; formerly Worcester Polytechnic Institute) specializing in Bayesian computation, Monte Carlo methods, survival analysis, and prior elicitation from historical data.

Key Contributions / Features

• Promotion-time cure-rate model (Chen-Ibrahim-Sinha 1999): A Bayesian cure model with a proportional-hazards structure, latent-variable Gibbs sampling, and valid noninformative-prior inference. Journal of the American Statistical Association 94(447): 909–919.
• Bayesian variable selection for logistic regression (Chen-Ibrahim-Yiannoutsos 1999): informative priors on coefficients (via a prediction $y_0$, precision $a_0$) and on the model space; posterior model probabilities from a single model's Gibbs output. Journal of the Royal Statistical Society, Series B 61(1): 223–242.
• Bayesian mortality modeling with random effects (Chen-Dey-Sinha 2000): time-dependent random-effects logistic (discretized) hazard model for toxicological multivariate mortality data with large families; historical-data power prior; variable selection and predictive model comparison. Applied Statistics 49(1): 129–144.
• Monte Carlo methods: Monte Carlo estimation of Bayesian quantities (Chen-Shao); Monte Carlo Methods in Bayesian Computation (Chen-Shao-Ibrahim 2000).
• Power priors: informative priors from historical data (with Ibrahim).

Related

• Cure-Rate Model
• Power Prior
• Variable Selection
• Chen-Ibrahim-Yiannoutsos (1999)
• Constantin Yiannoutsos
• Data Augmentation
• Chen-Dey-Sinha (2000)
• Dipak K. Dey
• Joseph G. Ibrahim
• Debajyoti Sinha
• Chen-Ibrahim-Sinha (1999)