基于决策倾向度的样本过滤与主动选择 |
| |
引用本文: | 陈科,唐雪飞. 基于决策倾向度的样本过滤与主动选择[J]. 电子科技大学学报(自然科学版), 2019, 48(3): 427-431. DOI: 10.3969/j.issn.1001-0548.2019.03.019 |
| |
作者姓名: | 陈科 唐雪飞 |
| |
作者单位: | 四川大学锦城学院计算机与软件学院 成都 611731;电子科技大学信息与软件工程学院 成都 610054 |
| |
基金项目: | 四川省重点研发项目2017GZ0192 |
| |
摘 要: | 该文提出了基于过滤函数的粗糙集样本决策倾向度动态调节与主动选择方法。首先定义样本过滤函数,从而确定样本选择或丢弃的依据;然后依次添加新增样本,根据过滤函数决定样本的去留,同时根据阈值指标调节已有样本的决策倾向度;最终建立有效的决策样本库,并在此基础上进行属性约简。本方法克服了传统变精度方法实现过程复杂、计算量大的问题,可有效地去除噪声数据,提高系统的鲁棒性。数据实验结果表明,该方法可以有效地压缩数据,提高样本分析质量。
|
关 键 词: | 属性约简 决策倾向度 过滤函数 粗糙集 |
收稿时间: | 2018-05-21 |
Active Sample Selection Method Based on Decision Making Tendency |
| |
Affiliation: | 1.School of Computer and Software, Jincheng College of Sichuan University Chengdu 6117312.School of Information and Software Engineering, University of Electronic Science and Technology of China Chengdu 610054 |
| |
Abstract: | A dynamic adjustment and active selection method for rough set decision making based on filtering function is proposed. Firstly, a sample filtering function is defined to determine the basis for sample selection or discarding; then, new samples are added in turn to determine the retention of samples according to the filtering function, and the decision-making tendency of existing samples is adjusted according to the threshold; finally, new sample library is established and attribute reduction is carried out. This method overcomes the problems of complex implementation process and large amount of calculation in traditional variable precision methods, and can effectively remove noise data and improve the robustness of the system. Experimental results show that this method can effectively compress data and improve the quality of sample analysis. |
| |
Keywords: | |
本文献已被 万方数据 等数据库收录! |
| 点击此处可从《电子科技大学学报(自然科学版)》浏览原始摘要信息 |
|
点击此处可从《电子科技大学学报(自然科学版)》下载全文 |