SVDD ALGORITHM BASED ON NOISE COST FOR ANOMALY DETECTION IN HYPERSPECTRAL IMAGERY
Journal Title: Applied Computer Letters (ACL) - Year 2017, Vol 1, Issue 2
Abstract
Anomaly detection algorithm based on Support Vector Data Description (SVDD) brings low detection rates due to background training samples being contaminated by anomalous data. To solve the problem, a new method based on SVDD with Noise Cost is proposed by introducing unbalanced data mining cost sensitive mind. This algorithm gives a different noise cost value to each background training samples through the neighbourhood clustering and then introduces the noise cost into SVDD to construct the SVDD hypersphere, thus making the classification interface more compact and improving the description ability of the anomaly and background value. At the same time, the sensitivity to the abnormal algorithm and the detection probability of the algorithm are greatly improved. Experimental results based on simulation data show that: compared to SVDD, this algorithm greatly reduces the false alarm rate, and improves the detection precision.
Authors and Affiliations
Liyan Zhang, Zhilin Liang, Xianling Zeng, Dandan Fu
Keywords
Related Articles
- EP ID EP403958
- DOI 10.26480/acl.02.2017.16.19
- Views 98
- Downloads 0
THE APPLICATION OF UNILATERAL SINGLE VALUE CONTROL CHART BASED ON LOGNORMAL DISTRIBUTION
The risk theory tells us that the short-term high risks exist at the right tail of the distribution. So we advocate the design and parameter estimation method, which the special figures that reflect short-term risks are...
RESEARCH ON APPLICATION OF DATA MINING BASED ON BAYESIAN NETWORK
This paper is committed to applied research in Bayesian networks in data mining, for there is not a complete Bayesian networks in data mining algorithms, propose a heuristic in data mining to construct Bayesian network u...
DYNAMIC RELIABILITY OF TORPEDO STRUCTURE UNDER THE ARBITRARY RANDOM EXCITATION
Aimed at the correlation between the impact loads and planing forces of the tail on supercavitating torpedo structure, the pseudo excitation had been constructed about every eigenpair of power spectral density matrix. By...
COMPUTER RESEARCH OF THEORY LINE LOSS BASED ON GIS COMPONENT TECHNOLOGY
The theoretical line loss data of distribution network is an important economical and technical index in Grid corporation routine work. This paper in view of the problems of theoretical line loss calculation in data inte...
GAIT PLANNING FOR ONE-LEGGED ARTICULATED HOPPING ROBOT
A gait planning procedure for planar one-legged articulated hopping robot is proposed in this paper. Its major advantage is the capability to design the take-off velocities of the mass center of the robot in X and Z axes...