IAP 2008: Methods for Large-scale Statistical Computing in the Social, Behavioral & Health Sciences

This course will help researchers performing large or complex statistical analyses to identify and analyze computational problems and thus improve performance, accuracy and reliability. Topics will include: fundamentals of computer arithmetic; computing architecture and performance; statistical benchmarking; principles of performance tuning; timing and profiling statistical codes; large database management; high-performance libraries; and distributed computing approaches.

The course is offered in a one-day mixed format. The morning portion of the class will be devoted to lecture and discussion. During the afternoon, the instructor will be available to offer one-on-one consulting on projects in either the planning or active stages. Please contact the instructor in advance to reserve a specific afternoon consulting time slot.

WHEN: Monday, January 28, 10 am – 1 pm & 1 – 4 pm (Individual consulting)

WHERE: E53-220

Please note that advance sign-up is required and participation is limited to 20 participants. We require prior familiarity with fundamentals of statistical model estimation.

Contact Micah Altman, Senior Research Scientist Institute for Quantitative Social Science, Harvard University, to sign up or with any questions.