The Biometrics Section is pleased to sponsor the following continue education courses at the 2018 Joint Statistical Meetings in Vancouver.
- Prediction in Event-Based Clinical Trials (half day),
- Regression Modeling Strategies (full day)
- Statistical Machine Learning for Biomedical Data (full day)
- Introduction to Bayesian Nonparametric Methods for Causal Inference (full day)
- Joint Modeling of Longitudinal and Time-to-Event Data (full day)
- Data Science for Statisticians (half day)
- Health Care Analytics in the Presence of Big Data (half day).
Introductions of the CE courses are listed below:
Prediction in Event-Based Clinical Trials
Instructors: Daniel Heitjan & Gui-shuang Ying
Did you ever wish you could use the accumulating data from your event-based clinical trial to reliably predict its future course? Well, now you can! Give these instructors a half day at JSM 2018 and they will teach you how - using their Bayesian simulation methods coded in straightforward R. Participants will learn about flexible parametric and nonparametric prediction models for simulating future enrollment and event histories. They will describe applications to real trials, showing how you can predict the timing of future interim analyses, identify efficient enrollment strategies informed by current data, and give DSMBs the best possible information on the likelihood of trial success. Bring your own computer and data and give their methods a try!
Health Care Analytics in the Presence of Big Data
Instructor: Evan Carey
The phrase "big data" has become widespread, but what does it mean for the practicing healthcare analyst? Come to this course to learn more! In this course, participants will gain hands-on experience using cutting edge software tools for the analysis of large administrative healthcare datasets, with a focus on Python and Apache Spark. Serial and parallel optimizations techniques using frequentist statistical frameworks and machine learning frameworks will be demonstrated. This course will focus on methods and software rather than the clinical context, but numerous real-world examples will be discussed which will offer a broad perspective. Students will be provided with a copy of a functioning "virtual machine" with all software and course materials pre-installed.
Regression Modeling Strategies
Instructor: Frank Harrell
When was the last time you had a "statistical modeling tune-up"? How do you keep up to date with methods for developing and validating predictive models, dealing with common analytical challenges, and graphically interpreting regression models? This course is the answer! Here is an enlightening and extremely popular course (that's why we offer it nearly every year) which covers multivariable regression modeling strategies, relaxing linearity assumptions, interaction surfaces, differences with machine learning, classification vs. prediction, quantifying predictive accuracy, detailed case studies using R, and more.
Introduction to Bayesian Nonparametric Methods for Causal Inference
Instructors: Jason Roy & Michael Daniels
Have you ever thought about trying more innovative approaches to causal inference, but you didn't know how to begin? Bayesian nonparametric methods (BNP) could be exactly what you are looking for! In this short course, these experts will review BNP methods and illustrate their use for causal inference in the setting of point treatments, dynamic (longitudinal) treatments, and mediation. The BNP approach to causal inference has several possible advantages over popular semiparametric methods, including efficiency gains, the ease of causal inference on any functionals of the distribution of potential outcomes, the use of prior information, and capturing uncertainty about causal assumption via informative prior distributions. You'll learn even more from their wealth of examples, supported by detailed instructions on software implementation using R.
Joint Modeling of Longitudinal and Time-to-Event Data
Instructors: Gang Li, Robert Elashoff & Ning Li
Have you ever needed to analyze a longitudinal study, but not sure how to handle non-ignorable missing data, informative visit times, or intermittently measured time-dependent covariates? Joint modeling methods could be exactly what you are looking for! In this short course, these experts will review the state-of- the-art statistical methodology developed in recent years for joint modeling of longitudinal and time-to-event data. They will provide motivating examples and an overview of statistical modeling and concepts that are fundamental to understanding joint models, and discuss several main areas in which joint models have been developed to address important scientific questions and challenging statistical issues. You will also gain hands-on experience of using joint models to analyze longitudinal and survival data, supported by detailed instructions on software implementation using R.
Data Science for Statisticians
Instructor: Rafael Irizarry
Have you been asked to teach a data science course and need guidance on what to teach? Come to this course and find out what we have learned from several years of experience teaching an introductory data science course. Demand for data science education is surging and traditional courses, offered by statistics departments, are not meeting the needs of those seeking training. A popular recommendation for improvement is that computing should play a more prominent role. We agree with this recommendation, but also advocate that the main priority is to bring applications to the forefront. In this short course we will work through some real world data analysis examples and, in the process, describe how we introduce skills and concepts not typically taught in traditional courses. Examples include data wrangling, exploratory data analysis, data visualization, reproducible research, and machine learning. We will also introduce tidyverse tools such as dplyr and ggplot2 which we find to be more effective for teaching beginners.
Statistical Machine Learning for Biomedical Data
Instructor: Noah Simon
What are the most common pitfalls in supervised learning? Do you know how they can be avoided? Come to this course and sharpen your skills! In this short course, participants will learn the "ins and outs" of supervised learning with Big Biomedical data. The course will cover modern statistical learning tools, illustrated using various biomedical examples (including examples in precision medicine). In particular this course covers penalized approaches to regression and classification; as well as support vector machines, and tree-based methods with emphasis on the analysis of "high-dimensional Omics" data sets. Each topic will be illustrated with examples, both of well-done and poorly-done analyses. This course is appropriate for students seeking an accelerated introduction to statistical machine learning or those who wish to better understand the nuances of these ideas and tools.