Spring Course 2014-15: Statistical Learning from Omics Data

This course will provide a broad overview of statistical methods for omics-type data through a multi-thread approach. The one thread will focus on statistical methods for so-called "large p, small n" data ranging from independent screening methods to regularization approaches including both Bayesian and frequentist methods. An overview of the variety of approaches being used, the details involved in their implementation and their advantages and disadvantages will be discussed. In additional, motivation will be provided for ongoing research topics of interest and under-studied areas. The second primary thread will focus on particular types of omics data applications, providing an overview of the study design, data structure, possible artifacts, scientific interests, current standard methods and ongoing problems. These application areas will include genomewide association studies, high-throughput sequencing, copy number variation, and emerging areas such as connectomics.

Pre-requisites:
The students should be at the level of a first year PhD student in statistics or a closely-related field, having a familiarity with calculus, linear algebra, basic probability theory, and basic statistics including maximum likelihood, simple hypothesis testing and linear regression.

Format:
This will be primarily a lecture style course with the grade being based primarily on attendance, class participation, quizzes having conceptual problems, and a course project.

Registration for this course is being processed through your respective university:
UNC-CH: STOR 892
Duke: 790.01
NCSU: 810.002

For additional information about this course, send e-mail to [email protected]