Challenges in Business Analytics; an industrial research perspective. 

Abstract: The Mathematical Sciences group at IBM Research is involved 
in a number of internal and client external projects.  In this tutorial 
we will present some of these projects and describe the challenges and 
issues we encounter in an environment where information technology 
becomes more prevalent.  With the greater availability of data and the 
sophistication of clients, there is a growing need to tackle aspects 
not only from a mathematical perspective, but also from a computer 
science perspective.  Data analysis must now consider aspects of data 
models, software architecture, data security, dynamic learning, just to 
name a few.  Our optimization models need to take advantage of this 
abundance of data, incorporate stochastic aspects, be adaptive and 
dynamic and deal with data uncertainty.  Furthermore, we see a growing 
client willingness to use mathematical models to address business problems.  
However, it is not always easy to understand and translate their needs 
into our models and often the inability to articulate our capabilities 
in a business language can hinder a project and produce negative results.   

Eleni Pratsini is the head of the Mathematical and Computational Sciences
Department at IBM Zürich Research Laboratories, Switzerland.