Decision Optimization

Decision Optimization

Delivers prescriptive analytics capabilities and decision intelligence to improve decision-making.

 View Only
Back to discussions
Expand all | Collapse all

Cut parameters and Optimality

  • 1.  Cut parameters and Optimality

    Posted Sat May 14, 2016 08:38 AM

    Originally posted by: Rym


    Hi

     

    I work on ILP models(binary) and often i fall on a NP-hard instance. So , according to your suggestions , i'm cranking up all cut parameters to obtain the optimal solution.
     
    But, i observe that there are a penality  of "Solving time":the time of solving is very big. Any suggestions for the modification of cut parametres in odre to minimize the solving time and saving the obtained solution?

    This is the modified cut parameters:

    execute
     
    {cplex.workmem=5000.0;
    cplex.mipemphasis=2;
    cplex.aggcutlim=6;
    cplex.cliques=3;
    cplex.covers=3;
    cplex.disjcuts=3;
    cplex.flowcovers=2;
    cplex.fraccuts=2;
    cplex.gubcovers=2;
    cplex.implbd=2;
    cplex.mircuts=2;
    cplex.mcfcuts=2;
    cplex.flowpaths=2;
    cplex.localimplbd=3;
    cplex.bqpcuts=3;
     
     
      }

     

    The model,data  and log files are attached.

     
    Thanks

    #CPLEXOptimizers
    #DecisionOptimization


  • 2.  Re: Cut parameters and Optimality

    Posted Mon May 23, 2016 02:14 AM

    According to the log a lot of time is burnt in disjunctive cut separation. So one thing to try would be to reduce the aggressiveness of those (or even disable them).


    #CPLEXOptimizers
    #DecisionOptimization


Related Content