One To One Marketing Optimization Algorithm The Benchmark Verification Method Study

Because of the restriction of time and memory, current optimization algorithm can't solve the large-scale One to One marketing optimization problem.However, we can develop a new optimization algorithm in the foundation of current typical optimization algorithm,this demand bring upping a kind of verification method to verify the performance of these optimization algorithm while begging solve the One to One marketing optimization problem.Therefore, this paper proposed a BenchMark method that will integrate theoretical analysis with practice function.Using this method , ten typical optimization algorithm have been exercised BenchMark which was based on theory and advanced language in part, and use the benchmark results we figured out that the LIPSOL method is the best one to be developed into One to One optimization algorithm. In the course of improving, the new optimization algorithm has been exercised BenchMark constantly to validate whether the new optimization algorithm has been improved, whether it has reached the anticipate effect. At the same time, the BenchMark result also figured out the direction to futher improve the algorithm, So we can have a definite object in view to improving the optimization algorithm and as the result the new algorithm has been improved gradually. Finally, The paper choiced the software GLPK which is a popular optimization software to excise BenchMark validation with the new One to One optimization algorithm, the BenchMark result indicated the advantage of WHCRM.In order to carry out BenchMark successfully, this paper established a set of One to One marketing optimization model, and in view of this model, we produced the BenchMark testing data based on Database and MPS format partly. These works established the firm foundation of BenchMark.