KDDAKernel Direct Discriminant Analysis
KDDAKennesaw Downtown Development Authority (Georgia)
KDDAKnowledge Discovery and Data Analysis
References in periodicals archive ?
for KDDA and LDA, the eigenvalue is proportional to the discriminative information quantity of each component, whereas for LDE, the eigenvalue is proportional to the ratio of the local between-class distance to the local within-class distance of neighbor samples.
Despite the higher accuracy, KDDA requires considerably more computation cost than the other methods because all M samples in the target region are involved.
two real-time RSS samples), our method and KDDA already outperform the other methods using 5 s, indicating that we can save 4 s for each positioning procedure while achieving higher positioning accuracy.