Global AI and Data Science

 View Only
Expand all | Collapse all

Emmanuel Katt Dubai | Optimizing Large-Scale Transportation Models with IBM Decision Optimization

  • 1.  Emmanuel Katt Dubai | Optimizing Large-Scale Transportation Models with IBM Decision Optimization

    Posted 16 days ago

    Hi everyone,

    I'm Emmanuel Katto from Dubai, United Arab Emirates (UAE), currently tackling a large-scale transportation optimization model and am running into some challenges with computational efficiency. Given the complexity of the problem, I'm wondering if anyone can recommend specific algorithms or tools from IBM's Decision Optimization portfolio that are well-suited for handling such large-scale models.

    Additionally, I'd love to hear about any strategies or best practices you've used in similar optimization scenarios, especially when dealing with complex, real-world transportation problems.

    Looking forward to hearing your suggestions and experiences!

    Thanks in advance,

    Emmanuel Katto 



    ------------------------------
    Emmanuel Katto
    ------------------------------


  • 2.  RE: Emmanuel Katt Dubai | Optimizing Large-Scale Transportation Models with IBM Decision Optimization

    Posted 16 days ago

    Presumably, you are using IBM ILOG CPLEX.  There are lots of different algorithms and parameters that affect CPLEX performance.  However, one of the keys to getting good performance from CPLEX or any other optimization solver is to use good modeling techniques.  Good modeling is both an art and a science.

    Our firm offers a Model Review and Validation Service that has helped our customers improve the performance of optimizers on their specific optimization problems.  We do so by improving the model, less by playing with specific algorithms or solver parameters.

    So my suggestion is that you engage a consultant to help you with your performance issues.



    ------------------------------
    Irv Lustig
    Certified Analytics Professional
    INFORMS Fellow
    Optimization Principal
    Princeton Consultants
    ------------------------------