My dissertation deals with several distinct concepts, including Bayesian computational methods, model selection in network meta-analysis with clinical trial data, and quantifying forecast uncertainty in machine learning.

Highlights:

• The first chapter serves as a “one-stop shop” for bayesian computational methods. Common MCMC methods are studied, and a simulation study compares the performance of several popular Bayesian computing platforms in R.
• The second chapter we assesses the performance of several model selection criteria for a complex family of network meta- analysis models for survival data.
• We also propose a technique for study outlier detection and present simulation results that demonstrate its effectiveness. Finally, the third chap- ter covers methods for constructing prediction intervals for forecasts for various machine learning algorithms.
• The third chapter proposes a new technique for measuring forecast uncertainty that can be used on a wide set of machine learning models.

I presented a talk on common Bayesian computational methods and implementations, which contrasts each and gives general guidelines for use. The paper form of this presentation is being written and will be a chapter in my dissertation.

Highlights:

• Gives overview of Gibbs Sampling and Hamiltonian Monte Carlo methods.
• The presentation addresses strengths and weaknesses of STAN, JAGS, OpenBUGS, Nimble, and Greta, and compares their respective computation time and accuracy.