Research on artificial immune algorithm based on controllable optimal objectives

We investigated several existing artificial immune models and there are not involve object controlled function and possess a memory network with dynamic change. The paper proposed a clustering algorithm of artificial immune network based on controllable optimal objectives. In the algorithm, the compression and clustering are abstracted as a multi-objective planning problem. The learning ability of immune system is improved by adopting the pool of memory cells strategy. The simulation of kernel clustering shows a satisfying result can be acquired by using the immune model with controllable optimal objectives.