John Hopkins University recently started a Data Science Specialization track consisting of 9 courses covering various topics such as R, statistical, and regression models. One of the courses focuses on reproducible research which is the idea that data analyses, and more generally, scientific claims, are published with their data and software code so that others may verify the findings and build upon them. For May's event, we are honored to have one of the instructors of that course, Dr. Roger Peng, come discuss reproducible research with evidence-based data analysis.
6:30 PM -- Networking & Food
7:00 PM -- Greetings
7:05 PM -- Reproducible Research with Evidence-based Data Analysis - Dr. Roger Peng
8:30 PM -- Post-event Drinks
Google maps does a good job of pinpointing the exact building location based on the address above.
Parking is not an issue. You can park on the street nearby or you can take up any Visitor spot that you can find. Also, feel free to park in the Ad.com/AOL employee parking lot but not in spaces marked Under Armour or Reserved.
Reproducible Research with Evidence-based Data Analysis
Statistical software is plentiful today, with new procedures andalgorithms constantly being developed, implemented, and optimized.Traditional statistical software tends to focus on solving arelatively self-contained task, often something that is a single pieceof a much larger data analysis. Data analysts are subsequently free tocombine the various pieces of statistical software out there in anynumber of combinations to analyze their data as they see fit. Hence, the number of "degrees of freedom" given to the analyst in most situations is enormous. But why is this so? Statistical software is typically written with a specific interface where certain parameters are modifiable by the user but most others are not. In this talk, Dr. Peng will discuss how a similar approach needs to be taken at the much higher level of the entire data analysis. Dr. Peng will introduce analysis pipelines---"transparent boxes"---that would have relatively few options available to the user and would be deterministic in their operation.
Roger D. Peng is an Associate Professor of Biostatistics at the Johns Hopkins Bloomberg School of Public Health and a Co-Editor of the Simply Statistics blog. He received his Ph.D. in Statistics from the University of California, Los Angeles and is a prominent researcher in the areas of air pollution and health risk assessment and statistical methods for environmental data. He created the course Statistical Programming at Johns Hopkins as a way to introduce students to the computational tools for data analysis. Dr. Peng is also a national leader in the area of methods and standards for reproducible research and is the Reproducible Research editor for the journal Biostatistics. His research is highly interdisciplinary and his work has been published in major substantive and statistical journals, including the Journal of the American Medical Association and the Journal of the Royal Statistical Society. Dr. Peng is the author of more than a dozen software packages implementing statistical methods for environmental studies, methods for reproducible research, and data distribution tools. He has also given workshops, tutorials, and short courses in statistical computing and data analysis.
Johns Hopkins Bloomberg School of Public Health is a leading international authority on the improvement of health and prevention of disease and disability. The school's mission is to protect populations from illness and injury by pioneering new research, deploying its knowledge and expertise in the field, and educating scientists and practitioners in the global defense of human life.
1606 Whetstone Way
Baltimore, MD 21230