Propositions de theses IBM

Bonjour a tous,
voici quelques propositions de theses a IBM France plus ou moins en lien avec les contraintes.

IBM France Lab – leader in the development of Decision Automation and Optimization products – is opening four Cifre PhD positions,

• Assembling and configuring existing optimization models to ease the creation of new solutions out of previous ones. At the crossing of Mathematical Modelling, Ontology reasoning and Software Product Lines configuration, this ongoing development needs further (applied) research to reach its full potential.

• Integrating Bayesian reasoning and classical rules-based reasoning. This approach extends the power of existing decision automation techniques and tools to new applications where yes/no decision cannot be achieved. Some experiments have already been developed; we want to jump to the next stage.

• Explaining automated decisions. This problem is still quite open though it is becoming a key request from user of automated decision systems. Accurate explanation might be of various forms, depending on their objective – audit trails or cognitive support of users – and on the level of knowledge of their users.

• Creating Adaptive Decision Models: decision models are used and managed by regular business users, without the need of I.T. specialists. We are shooting at the next step: let the models evolve smoothly accordingly to their objectives. Analytics tools can be put in place that use past decisions to produce recommendations or adaptations of the existing models to better fit the objectives.

IBM France Lab is the IBM R&D Lab based in France – located in Paris, Sophia- Antipolis, Toulouse and Pornichet. It hosts a global team of about 700 developers and high profile technical people. Our core activity is the development of leading software products that are delivered and used by customers all over the globe. The open PhD positions are related to two very successful products lines :

• Decision Management: generalisation of the ‘Business Rules’ approach, now widely used to automate operational decisions represented by set of dozens, hundreds or thousands rules and decision tables. France Lab is proud to be the world-wide leader in Decision Management.

• Decision Optimization : these algorithms and modelling languages enable taking the best decision when using limited resources. Thanks to its ILOG CPLEX and ODME product offerings IBM is the leader in the field.

France lab is looking for brilliant and dedicated individuals passionate for Decision Automation and Optimization techniques and languages. The PhD Candidate will be co-supervised by a university professor and an expert from the Lab, he or she will be fluent in English, have good skills in mathematics as well as in programming. Ideally candidates have a solid background in Artificial Intelligence and/or Operations Research.

Candidates email your resume to or