Applied Statistics and Data Science

Course Description

PART I: Elements of a scientific problem, including estimands, the role of statistical models, the language of statistical modeling, notions of likelihood, finite vs infinite populations, and types of analysis. PART II: Elements of statistical modeling, including transformation theorems, sufficiency, 1-parameter and multi-parameter models, multivariate Normal models, Dirichlet-multinomial models, hierarchical models, generalized linear models, mixture models, text analysis, social network analysis. PART III: Concepts and algorithms for estimation and inference, including information, statistical efficiency, asymptotic approximations, maximum likelihood estimators, method of moments estimators, Bayesian estimators, empirical Bayes vs full Bayes estimation strategies, expectation-maximization algorithm, Monte Carlo approximations, Gibbs samplers, Metropolis-Hastings samplers, prior and posterior predictive checks, and Bayesian vs. frequentist coverage. Data Science visitors: The course will feature a series of short talks and Q&A sessions with prominent data scientists spanning academia, government, and the Tech industry.

Code

STAT 8109

Credits

3

Subject

STAT

Programs That Include This Course

• Doctoral Doctor of Philosophy in Statistics
• Graduate Certificate Graduate Certificate in Business Analytics
• Master's Master of Science in Business Analytics
• Master's Master of Science in Statistics and Data Science