针对目前指纹室内定位系统指纹库管理效率低、实时性差和定位精度低的问题,提出了一种新的基于核化K-means和SVM分类回归的无线定位算法。首先利用核化K-means算法将输入的预处理后的RSS(Received Signal Strength)信号进行无监督聚类,...针对目前指纹室内定位系统指纹库管理效率低、实时性差和定位精度低的问题,提出了一种新的基于核化K-means和SVM分类回归的无线定位算法。首先利用核化K-means算法将输入的预处理后的RSS(Received Signal Strength)信号进行无监督聚类,将聚类后的数据信息存入指纹特征数据库,然后通过SVM回归的机器学习算法对特征数据库的数据进行训练,得到一种最优的拟合位置函数的数学模型。并且采用粒子群算法对参数进行寻优,进行实验仿真。实验结果表明,该算法有效地提升了定位精度,优于KNN、WKNN、SVR等室内定位算法。展开更多
In this paper, we present an adaptive anomaly detection framework that isapplicable to network-based intrusion detection. Our framework employs fuzzy cluster algorithm to detect anomalies in an online, adaptive fashio...In this paper, we present an adaptive anomaly detection framework that isapplicable to network-based intrusion detection. Our framework employs fuzzy cluster algorithm to detect anomalies in an online, adaptive fashion without a priori knowledge of the underlying data. We evaluate our method by performing experiments over network records from the KDD CUP99 data set.展开更多
文摘针对目前指纹室内定位系统指纹库管理效率低、实时性差和定位精度低的问题,提出了一种新的基于核化K-means和SVM分类回归的无线定位算法。首先利用核化K-means算法将输入的预处理后的RSS(Received Signal Strength)信号进行无监督聚类,将聚类后的数据信息存入指纹特征数据库,然后通过SVM回归的机器学习算法对特征数据库的数据进行训练,得到一种最优的拟合位置函数的数学模型。并且采用粒子群算法对参数进行寻优,进行实验仿真。实验结果表明,该算法有效地提升了定位精度,优于KNN、WKNN、SVR等室内定位算法。
基金广东省自然科学基金( the Natural Science Foundation of Guangdong Province of China under Grant No.04300504) 广东省科技攻关项目( the Key Technologies R&D Program of Guangdong( Province) +2 种基金 China under Grant No.2005B20701008 No.2005B10101028 No.2004B20701006)。
基金Supported by the National Natural Science Foun-dation of China (60573101) the Natural Science Foundation ofShaanxi Province (2005f43)
文摘In this paper, we present an adaptive anomaly detection framework that isapplicable to network-based intrusion detection. Our framework employs fuzzy cluster algorithm to detect anomalies in an online, adaptive fashion without a priori knowledge of the underlying data. We evaluate our method by performing experiments over network records from the KDD CUP99 data set.
