Presse
Wall Street Journal (European Edition), 25. Juli 2013
... Professors Brieden and Gritzmann have already successfully developed extended and specifically customized models for several problem areas, including insurance loss prediction, maintenance prediction, and shipping logistics. The foundations of their algorithms are deep geometric insights that lead to high-quality segmentation of big data. Part of what makes their work particularly interesting for practitioners is the fact that it is a non-black-box algorithm which works with small setup costs and has high predictive power. "This groundbreaking work shows how mathematics can drive innovation in areas all across business and society," said Oliver Bastert, product manager for FICO's optimization software. "As just one example, Gritzmann and Brieden developed a method of insurance loss forecasting that overcomes the weaknesses of today's standard method, GLM, while simultaneously delivering a significant uplift in profitability." ...
s. auch