### Table 4. Kernels and Applications

2003

"... In PAGE 9: ... 5. Performance Evaluation Performance was evaluated with six media processing kernels and applications, summarized in Table4 . Kernels and applications were written in KernelC and StreamC.... ..."

Cited by 18

### Table 4. Kernels and Applications

2003

"... In PAGE 9: ... 5. Performance Evaluation Performance was evaluated with six media processing kernels and applications, summarized in Table4 . Kernels and applications were written in KernelC and StreamC.... ..."

Cited by 18

### Table A.11 Kernel

1994

Cited by 86

### Table 1: Kernel Notation

2005

"... In PAGE 2: ... (Cristianini 2001) Composite Kernels Proving that a function is a kernel is relatively difficult. However, there are several well known kernels, including the polynomial and radial basis function (RBF) kernels, shown in Table1 . We leverage these established kernel functions by applying operators that are closed over the kernel property (Joachims, Cristianini, amp; Shawe-Taylor 2001).... In PAGE 2: ... A table of kernel composition operators and the following theorem appear in (Joachims, Cristianini, amp; Shawe-Taylor 2001). With this information, the four basic kernels shown in Table1 can be used to create an infinite combination of composite kernels. Theorem 2 (Composite Kernels).... In PAGE 4: ... A hill climb- ing search with random restart was used to enumerate com- posite kernel parameters, and the model with the best per- formance estimator was selected. A scanner-parser was developed that recognizes the ker- nels shown in Table1 and the composition rules described in Theorem 2. An interpreter was developed to evaluate the value of the parsed kernel function.... ..."

### Table 1: Kernel Notation

"... In PAGE 2: ... (Cristianini 2001) Composite Kernels Proving that a function is a kernel is relatively difficult. However, there are several well known kernels, including the polynomial and radial basis function (RBF) kernels, shown in Table1 . We leverage these established kernel functions by applying operators that are closed over the kernel property (Joachims, Cristianini, amp; Shawe-Taylor 2001).... In PAGE 2: ... A table of kernel composition operators and the following theorem appear in (Joachims, Cristianini, amp; Shawe-Taylor 2001). With this information, the four basic kernels shown in Table1 can be used to create an infinite combination of composite kernels. Theorem 2 (Composite Kernels).... In PAGE 4: ... A hill climb- ing search with random restart was used to enumerate com- posite kernel parameters, and the model with the best per- formance estimator was selected. A scanner-parser was developed that recognizes the ker- nels shown in Table1 and the composition rules described in Theorem 2. An interpreter was developed to evaluate the value of the parsed kernel function.... ..."

### Table 1. Common kernels

2002

"... In PAGE 2: ... Optimizing the SVM hyper-parameters is a model selection problem that needs adapting multiple parameter values at the same time. The parameters to tune are those that embed any kernel function as the parameter in an RBF kernel or the couple ( ; ) in case of KMOD kernel (see Table1 ). In addition, another parameter the optimization may consider is the trade-off parameter C which may have a strong effect on the SVM behavior for hard classification tasks.... ..."

Cited by 4

### Table 2. Kernel Mutability.

1997

"... In PAGE 3: ... The first two questions determine the mutability of the system while the third indicates the location of extensions, implying the types of protection mechanisms that might be employed to protect the kernel from errant extensions. Table2 shows the cross-product of mutability options for a system, providing examples of systems using each approach. It is important to note that a single system might employ multiple approaches to extensibility and can implement its approach either at user-level or at kernel-level.... ..."

Cited by 12

### Table 4: Kernel functions

"... In PAGE 11: ... Instead, we restrict attention to mappings for which the kernel K(xi; xj) = h (xi); (xj)i may be computed e ciently, without rst mapping the patterns to RD. The formulae for several common kernels are shown in Table4 . We used this property in our experiments with the SVM.... ..."

### Table 1. DSPstone Kernels

"... In PAGE 10: ... These three cases are referred to as X-Allocating, Scheduling,andPartitioning, respectively. Table1 lists the performance results obtained for some selected DSPstone kernels. Each kernel contains some loops with operations on two or three global arrays.... ..."

### Table 1: Kernels

2005

"... In PAGE 4: ...ver various runs. The variation is not signi cant. 3.6 Detailed Analysis Table1 presents the top kernel functions for FPGrowth, Genmax, and Apriori. In FPGrowth, 61% of the exe- cution time is spent in the Count()-FPGrowth routine.... ..."

Cited by 13