【摘要】 在核爆地震模式识别中,特征选择既是非常关键同时又是非常困难的一步,为解决该问题,应用序优化方法对此进行了研究,并利用序优化中的盲选规则(BP)进行了特征选择试验。通过将设定的2组匹配概率和实际获得的实验结果进行比较,发现序优化方法能在不需要任何先验知识的前提下迅速地减小搜索空间,提高搜索效率。最后将盲选规则用于核爆炸和地震的特征选择中,取得了97.5%的识别率,这为核爆地震识别中的特征选择问题提供了一种新的解决方案。更多还原

【Abstract】 Feature selection is a very important and difficult problem in the pattern recognition of underground nuclear explosions and natural earthquakes. To solve this problem, a new method for feature selection based on ordinal optimization was presented in this paper, and an experiment based on blind picking rule was carried out. Experiment results and corresponding analysis indicate that this method can reduce the search space greatly and enhance the search efficiency. The blind picking rule is applied to the feature selection of underground nuclear explosions and natural earthquakes, and the recognition rate reaches up to 97.5%. It may be a novel solution for the feature selection of underground nuclear explosion and natural earthquake events.更多还原