Sentence examples for kernel diffusion from inspiring English sources

Application of kernel diffusion to a physical interaction network for a range of scales yields a graph topological scale-space.

To best capture the relationships among host resistance genes, we evaluated the performance of several different network similarity measures: direct interaction ranking (DIR, STRINGG association ranking (SAR), random walk with restart (RWR), and seed-based heat kernel diffusion ranking (sHKDR).

In particular, DNR applies equal weights for all neighbors; by comparison, deHKDR considers the initial interaction scores between the studied gene and its neighbors and applies the final weights from the heat kernel diffusion matrix.

Nonetheless, much more complex similarity coefficients [ 52] can be used as kernels on graphs (e.g., exponential diffusion kernel, Laplacian exponential diffusion kernel, or the commute time kernel).

Following Fouss et al. (2006), there exists several kernels on graphs, such as the Exponential Diffusion Kernel, the regularized Laplacian Kernel, the von Neumann Diffusion Kernel, and the Commute Time Kernel [37].

Other graph kernels such as the diffusion kernel [58] are less suitable since they require an extra parameter to be optimized.

These new interpolation kernels incorporate diffusion as well as remeshing.

H n + ⋯ If a graph Γ with a Laplacian L is considered, then exp(− θ L is called the diffusion kernel or heat kernel for graph Γ, where θ is a rate of diffusion [ 25].

We used polynomial kernels for categorical data, Gaussian radial basis function (RBF) kernels for continuous data and diffusion kernels for graphs as described previously [4,6].

The predictive ability of RKHS models with either a diffusion kernel or a Gaussian kernel was assessed by cross-validation.

Kondor and Lafferty [ 25] obtained promising results when the diffusion kernel was compared with several kernels in classification problems with a set of discrete predictors, and this kernel has been used in a microarray-based gene function prediction problem [ 38].