Local Similarity-Based Fuzzy Multiple Kernel One-Class Support Vector Machine

<div>The classification performance of MKOCSVM and FMKOCSVM_LS. 10% of Gaussian noise is added. The magenta solid line is the boundary. The blue dots are the target samples. The black crosses are the noises with Gaussian distribution. The number at bottom of the noise represents its membership.</div>

Complexity

fig5

Figure 5

Figure 5: Local Similarity-Based Fuzzy Multiple Kernel One-Class Support Vector Machine 