Identifying key variables that impact risk or explain behavior has always been a core challenge in finance. In the arms race to do this at exponentially decreasing amounts of time for exponentially increasing amounts of data, the emerging technology of quadratic unconstrained binary optimization is a powerful weapon. 

Whether you are an algorithmic trader trying to detect sparse signals in a large and noisy market, a credit underwriter trying to interpret a vast number of features, or a payments processor trying to identify bad actors – finding the simplest solutions with large numbers of variables and not a lot of training samples is a hard but crucial task. We will explain how QUBO works, why it is well-suited for the task, and why LightSolver can be your partner.