For the second year in a row DAS Research presents at the world’s most important machine learning conference NeurIPS (Conference on Neural Information Processing Systems).
This year our colleagues Sören Laue and Joachim Giesen demonstrate generic optimization for Machine Learning (GENO). Their work is groundbreaking in several respects. So far, it has been widely held that efficient generic optimization is impossible. GENO shows not only that it is possible, but also by orders of magnitude faster, simpler and more versatile than alternative methods.
For a quick overview check out the conference poster here: http://www.geno-project.org/posterNeurIPS.pdf