Frank owns a Ph.D. in “Sciences Appliquées” (Applied Sciences) from the Université Libre de Bruxelles (ULB) with a specialization in mathematical optimization. He is the author of numerous high-end mathematical softwares and libraries.
Frank is the main developper of the TIMi Datamining Suite. Before creating TIMi, Frank was the main researcher, developper, software architect, chief technological officer, project team leader in a large number of software projects in Business Intelligence and Datamining, first at the Artificial Intelligence Laboratory of the ULB (IRIDIA) and after at other consultancy companies like Business&Decision and Vadis. Amongst other things, he created the datamining software named RANK still used today by some other consultancy companies to construct predictive models for the BNP Paribas/FORTIS bank in the retail and B2B field.
Frank is also a scientific researcher in the field of datamining and optimization. Its ambition is to “create the softwares of tomorrow”. He posses a deep expertise in these fields:
- Business Intelligence
- Analytical CRM & Predictive Techniques
- Linear Algebras: matrix factorizations, LAPACK, numerical stability of algorithms, LARS, Elastic Net, large scale computations, …
- Datamining: techniques to prevent overfitting such as cross Validation, Prediction techniques such as multinomial predictors, Decision trees, bootstraping techniques, linear identification technique…
- Optimization: Newton’s based algorithm, SQP, Trust Region Methods, LP, IP,…
- Scientific programmation
- High-efficiency computation (vectorized, parallel and disributed computing – optimized branching)
- I/O efficient semi-external algorithms
- Software Architect for large scientific softwares and libraries
- Low-level assembly language
- He is the author of the mini-SSE-L1-BLAS library
- Memory constrained & embedded programming
Frank’s knowledge in fields allows him to seek out new elegant and efficient solutions to almost any statistical and Business Intelligence problems.