Utilization of the Cloud in Matlab Analytics: A Case from the Business Humanities Research

Description

My research at Copenhagen Business School (CBS), “Computing Complexity of Cultures” employs highly advanced mathematical models embodied by the so-called “nonparametric Bayesian relational modeling” developed by the DTU Compute research group (Associate Professors, Morten Mørup and Mikkel N. Schmidt: brainconnectivity.compute.dtu.dk) at the Technical University of Denmark. This has been successfully applied to analyze, amongst others, complicated social networks and advanced brain connectivity data during the past few years. In this context my ambitious challenge is to apply these highly complex models to the new research field of cross-cultural data analysis. The mathematical models are expected to extract, visualize and map-out heterogeneous structures of human values/attitudes/behaviors across cultures, which has been considered highly challenging for more conventional cross-cultural data analysis tools. The cloud-based high performance computational solution provided by Microsoft and Techila Technology has enabled our joint CBS-DTU R&D to pursue preliminary investigations of these challenges for analyzing heterogeneous structures of survey response patterns between e.g. Swedes and Japanese as example case. Matlab codes developed by the DTU research group were directly executed on the Azure-Techila server, which has enabled the optimization of computational time to be substantially faster than what has been previously possible using the traditional computational environment and infrastructure available at a research institute like CBS, inherently rooted in humanities and social sciences.

Period	7 Oct 2015
Event type	Conference
Location	Middelfart, Denmark