以二分法为基础,应用自动学习机原理建立一种动态的感官评估实验方法.评估者根据评估过程中获得的信息,动态地调整评估策略.该实验方法分为二方面内容:首先提出应用二分法以较少实验次数完成对样本的评估,其次是应用自动学习机原理优化二分法.仿真结果表明,在不影响实验结果精度的前提下,动态实验方法能够显著地减少评估实验次数.
In this paper, a method of dynamic experiment designs in sensory evaluation is presented that is based on dichotomy, and the principle of learning automata is applied in optimizing the dichotomy. The evaluator adjusts the strategies according to the information that is obtained in the process of evaluation. This experiment method includes two aspects, and the first content is to reduce the number of tests according to dichotomy, and the second content is how to optimize the method of dichotomy through the principle of learning automata. It's proved that the method can reduce the times of tests greatly by the result of simulation with a high precision of evaluation.