The Statistics Seminar speaker for Wednesday, October 2, 2019, is Matthias Kormaksson, an associate director statistical consultant in the Statistical Methods and Consulting group at Novartis Pharmaceuticals. He works on a variety of statistical learning problems in clinical trials, in particular on unsupervised clustering problems for identifying disease subtypes and supervised problems for response predictions. From 2012 to 2017 he worked as a research staff member and statistical consultant at IBM Research in Brazil where he worked on the application of spatial and spatio-temporal predictive models to problems arising in the area of natural resources. He received his Ph.D. in Statistics at Cornell University in 2010 followed by a postdoc in the Division of Public Health and Epidemiology at Weill Cornell Medical School, where he did research on statistical methods involving microarray and next generation sequencing data in cancer epigenetics.
Talk: Statistical Methodology and Advanced Exploratory Analytics at Novartis
Abstract: In a Financial Times article from September 25, 2017 Vas Narasimhan, CEO of Novartis, outlined the swell of enthusiasm over the use of digital technology in R&D at Novartis. In the classical drug development setting biostatisticians have tended to focus on techniques suited to the confirmatory phase III paradigm. However, the growing area of data science has placed greater focus on exploratory analyses, supervised/unsupervised learning and other concepts from both the Statistical and Machine Learning literature. In this talk we will discuss some of our early experiences with incorporating this expanded data science skillset to the current drug development environment. Specifically, the talk will focus on some of the analytical challenges that we are currently facing at Novartis and different approaches to embracing them. In particular, we will present two use cases from extensive in-house drug development programs.