QPRC 2016

Spatio-Temporal Analysis for System Management at IBM Research

Yasuo Amemiya

IBM Research

Abstract

IBM has been providing analytics-based solutions to various large-scale problems relevant for business, government, and society.  A goal of such a project is to manage a large physical system effectively based on analysis of various measurements taken over space and time.  Statistical analysis methods and ideas are essential part of the overall solution development.  In particular, new types of spatio-temporal analysis methods are needed.  In this talk, some of large system management projects at IBM Research are described, and the development of appropriate spatio-temporal analysis methods is discussed.

Keywords: Statistics research in industry, spatio-temporal analysis, business analytics, integration of physical models.