### Table 1 The function set for generating high order polynomials. Here, quot;[.] quot; represents the list of polynomial terms generated by ap- plying GP functions.

"... In PAGE 6: ... If we take 18 different 2nd order polynomial functions fi F s that take two arguments in the GP function set FGP (Nikolaev and Iba 2001), it is very easy to create various high order polynomials with the tree structure. Table1 records all these functions. The terminal set TGP contains only variables.... In PAGE 6: ...tructure. Table 1 records all these functions. The terminal set TGP contains only variables. FGP = f+; ; f1 F ; : : : ; f18 F g TGP = fx1; : : : ; xng In Table1 , xr and xs are the arguments of fi F . If the tree is (f1 F x2 (f6 F x1 x3)), then its corresponding list becomes [f1 F x2 [1; x1; x1x3]]=[1; x2; x1x2; x1x2x3].... ..."

### TABLE 2 Parallel Newton high-order interpolation

### Table 6.2: Sizes of the high-order raw models.

2004

### Table C.2: Registration results C.1 Local dissimilarities of the surfaces The algorithm assumes that the surfaces are related through a global a ne transfor- mation. Therefore, when given the task of registering two surfaces, the algorithm only captures the global transformation and does not take into account the local features of the surfaces. Moreover, it does not deform a surface to match the other surface. One con- sequence of this behavior is that the algorithm can not precisely register the CT and the Cyberware datasets coming from two di erent individuals, since the facial skin surfaces of di erent individuals are not related through a global transformation. The following example illustrates this shortcoming of the algorithm. S1 and S2 are transformed using the transformation parameters listed in Table C.1. The transformed surfaces are called S3 and S4 respectively. We now register S1 with S3 and S4 and compute the ISD values.

### TABLE II THE EXPERIMENTALLY DETERMINED PROBABILITY ON SELECTING AN ith ORDER AR-MODEL, BASED ON 10,000 ORDER OBSERVED SEQUENCES, GIVEN N SAMPLES OF TWO 2nd-ORDER AR-PROCESSES AS A FUNCTION OF THE PROBABILITY ON SELECTING A TO HIGH-ORDER. IN CASE OF INDISTINGUISHABLE MODELS WE HAVE AN A PRIORI PREFERENCE FOR LOW COMPLEXITY (ORDER) MODELS.

### TABLE II The experimentally determined probability on selecting an ith order AR-model, based on 10,000 order observed sequences, given N samples of two 2nd-order AR-processes as a function of the probability on selecting a to high-order. In case of indistinguishable models we have an a priori preference for low complexity (order) models.

### Table 3: Average message latencies for low-order interleaved (LOI) and high-order interleaved (HOI) memory.

in Effect of Virtual Channels and Memory Organization on Cache-Coherent Shared-Memory Multiprocessors

1996

"... In PAGE 14: ... Figure 3: Tra c pattern for MATMUL. Table3 . In case of high-order interleaved memory, the execution time shows a big improve- ment when the number of virtual channels are increased from 2 to 4.... ..."

Cited by 2

### Table 2: Surface Parameterization for business jet problem

1997

"... In PAGE 12: ... Six design variables were used whose associated mode shapes were combinations of 3 chordwise functions and 2 spanwise functions. The selected design variables result in a wing parameterization given by equation (36) and the functions fi and gj for this case are listed in Table2 . Note that the chordwise functions are given by a shear function (which is similar to a twist variable for small geometry perturbations), and two Hicks-Henne functions.... ..."

Cited by 29

### Table 5: The best Minkowski values. The best r parameter value for the Minkowski distance in each environment and each local planner for free to surface (top) and surface to surface connections (bottom).

1998

"... In PAGE 17: ... gets harder, the slope of the curve decreases. Table5 shows that most of the local planners reach their highest score for the surface to surface connections at r = 1:5. Our analysis of the graphs generally showed that if r = 1:5 was not the highest score, then it was the second best.... In PAGE 19: ... Note that in Table 4 it is clearly shown that the optimal s value gets higher as the environment gets harder. The Minkowski parameter graphs (Figures 26 and 27) and Table5 display two common trends in the alpha-puzzle environment. The free to surface connection score always reaches the highest value at either r = 2:5 or r = 3 while the highest value for surface to surface connections is around r = 1:5.... ..."

Cited by 59