水声目标自动识别技术作为实现武器装备智能化的核心技术之一，具有重要的军事意义。水声目标识别模型评估是目标识别技术研究中的关键环节，对促进识别技术发展起到了十分重要的作用。国内在自动目标识别模型评估方面，特别是在评估理论与方法上的研究还刚刚起步，远不能满足当前深入持续发展目标识别技术的迫切需求。根据信息熵的理论及其应用，构建了水声目标识别模型性能评价指标体系；借鉴信息论的方法，利用信息熵确定模型综合评估中各指标的权重系数，提出客观的水声目标识别模型综合评估方法。
As one of the core technologies to realize the intelligence of weapons and equipment, underwater acoustic target automatic recognition technology has important military significance. Underwater acoustic target recognition model evaluation is a key link in the research of target recognition technology, and it plays a very important role in promoting the development of recognition technology. Domestic researching on evaluation method of acoustic target recognition models, especially in the evaluation theory and method, has just started, and it is far from meeting the urgent needs of the current deep and continuous development of target recognition technology. Based on the theory and application of information entropy, a performance evaluation index system for underwater acoustic target recognition model is constructed; using information theory method, information entropy is used to determine the weight coefficient of each index in the model comprehensive evaluation, and an objective underwater acoustic target recognition model is proposed comprehensive evaluation method.
2021,43(6): 134-137 收稿日期：2020-06-20
DOI：10.3404/j.issn.1672-7649.2021.06.025
分类号：TB566
作者简介：齐振(1987-)，男，工程师，研究方向为声呐作战使用
参考文献：
[1] 程玉胜, 李智忠, 邱家兴. 水声目标识别[M]. 北京: 科学出版社.
[2] 何峻. 自动目标识别评估方法研究[D]. 长沙: 国防科学技术大学, 2009.
[3] 吴军. 数学之美[M]. 北京: 人民邮电出版社, 2018: 60−65.
[4] 李宏东译. 模式分类(第二版)[M]. 北京: 机械工业出版社, 2005: 320-322.
[5] 周志华. 机器学习[M]. 北京: 清华大学出版社, 2016: 23−28.
[6] 韩家炜, 坎伯. 数据挖掘: 概念与技术[M]. 北京: 机械工业出版社.
[7] HAND DJ, TILL RJ. A simple generalisation of the area under the Roc curve for multiple class classification problems[J]. Machine Learning, 2001, 45(2): 171–186
[8] FAWCETT T. Using rule sets to maximize Roc performance[C]//Proceedings 2001 IEEE International Conference on Data Mining, [S.l.]: Ieee, 2001: 131−138.
[9] FERRI C, FLACH P, HERNÁNDEZ-ORALLO J. Decision trees for ranking: effect of new smoothing methods, new splitting criteria and simple pruning methods[J]. Technical Report, DSIC 2003, 2003
[10] PROVOST F, FAWCETT T, KOHAVI R. The case against accuracy estimation while comparing induction algorithms[C]//ICML Conference. 1998.
[11] 秦锋, 罗慧, 程泽凯, 等. 一种新的基于AUC的多类分类评估方法[J]. 计算机工程与应用, 2008, 44(5): 194–196
[12] 秦锋, 杨帆, 程泽凯, 等. BO-AUC多类分类评估方法[J]. 计算机工程与应用, 2012, 48(5): 156–158
[13] 仇方道. 县城可持续发展综合评价研究[J]. 经济地理, 2003, 23(3): 319
