Skip to content

Regression Modeling Strategies and the R rms Package Short Course 2019

4 messages · Frank E Harrell Jr, Graeme Davidson, David Winsemius

Original
Frank E Harrell Jr
# 
*Regression Modeling Strategies Short Course 2019*

Frank E. Harrell, Jr., Ph.D., Professor

Department of Biostatistics, Vanderbilt University School of Medicine

fharrell.com ? ? @f2harrell



*May 14-17, 2019* With Optional R Workshop May 13

9:00am - 4:00pm

Alumni Hall

Vanderbilt University

Nashville Tennessee USA



See http://biostat.mc.vanderbilt.edu/RMSShortCourse2019 for details.



The course includes statistical methodology, case studies, and use of

the R rms package. ?Emphasis is on developing predictive models, model validation, and quantifying predictive accuracy, plus many more topics including navigating the choice of statistical models vs. machine learning.







Frank E Harrell Jr?????	Professor? ? ?	School of Medicine

	Department of Biostatistics?????	Vanderbilt University
Graeme Davidson
# 
Hi Frank, 

As part of the R community, you will be aware that the vast majority of knowledge regarding statistics such as linear modelling is online for free. What makes this course worthy of payment compared to freely available information and/or well structured fee paying courses such as DataCamp? 

All the best

Graeme R Davidson PhD

Data and Insight Analyst
Frank E Harrell Jr
# 
You'll need to see hundreds of past course evaluations to see why, but a main reason is the learning environment we provide, with hours of discussion among participants occuring during the 4-day course.  This includes topics such as how to collaborate with non-statisticians especially when explaining the results of complex statistical models.  We also have discussions during the lunches in the dining hall next to the lecture room.  Many of the topics I cover are set by the participants during the week.  And all of the course handouts are freely available for everyone.

Frank


________________________________
Frank E Harrell Jr      Professor       School of Medicine

        Department of Biostatistics             Vanderbilt University
David Winsemius
# 
Hi Graeme;


I took the course about ten years ago. I did so after getting a Masters 
in Epidemiology from the University of Washington and doing very well in 
all my stats courses and submitting my thesis work on solving regression 
problems with stratified sampling using bootstrap methods. So I think I 
probably had a much more solid grounding in regression methods than the 
average Datacamp customer. I found the course work and discussion very 
useful. I already had a copy of Harrell's RMS text and had read much of 
it before that class as well as applying several of the methods he 
illustrated. It covered topics of validity in inference and modeling of 
covariate functional relationships in much greater depth than I have 
ever seen in any of the online material I have reviewed in the last ten 
years. It was 4 days well spent and far cheaper than I could have gotten 
from a typical consultation.

The typical online course work demonstrates the regression machinery but 
very rarely gets into the issues of modeling splines or penalized 
methods. Model comparison and assessment of validity is often given 
cursory treatment. I continue to see questions on CrossValidated.com and 
StackOverflow that demonstrate that the bulk of the self-learners or 
distance learners have so far failed to acquire the knowledge base that 
vould be acquired during Frank's course. I would advise someone who has 
taken a Datacamp course in regression methods to take Frank's course as 
the next step to being "statisitcally educated".