### Table 2: The reward matrix

"... In PAGE 7: ... We will refer to a stable population norm as a stable strategy. The matrix of rewards per period to a focal individual, using a given row strat- egy against a population using a given column strategy, can now be determined from (1)-(3), (5) and (8), and is shown in Table2 . We will denoted this matrix by A, so that aIJ is the reward to an individual using strategy I against n individuals us- ing strategy J.... In PAGE 7: ... Note that the bene ts of monitoring to the focal individual are independent of the number of monitors if no one else is poaching, as indicated by the subscripted dots in columns 5 and 6 of the payoff matrix. A population strategy is stable if its diagonal element in Table2 is the largest in its column. In other words, strategy J is stable if aJJ exceeds aIJ for all I 6 = J; or equivalently, if the only non-positive term in column J of the population stability... In PAGE 11: ... The strategy NM that supports the agreement is a cooperative strategy when the reward to each individual from not hunting and monitoring exceeds the bene t to each individual from hunting (with the higher-value technology) but not moni- toring, or strategy HX. In other words, and in terms of the payoff matrix A de ned by Table2 , NM is a cooperative strategy if a55 exceeds both a22 and a44 or f ? c0 + B n + 1 gt; VH: (19) The higher the value of cH, or the lower the value of RH, the greater the sig- ni cance of whether NM is a cooperative strategy. In general, if a strategy is the only stable one, then it will ultimately emerge as the community norm; whereas if a second strategy is also stable, then the rst will emerge only if it yields a higher com- munity reward.... ..."

### Table 1 shows the distance measure between face images as a matrix. We have used pairs of images from 10 individuals not displaying any emotion, that have been taken at times more than two weeks apart. We can show that corresponding face images are recognised in most of the cases as more similar than images from di erent individuals. Indeed, one can notice that for every line and every column, the diagonal element is the

"... In PAGE 8: ...no error is made. person no 1 2 3 4 5 6 7 8 9 10 1 1020 1410 1635 1430 1360 1513 1570 1516 1526 1530 2 1587 1034 1574 1587 1315 1309 1417 1459 1377 1403 3 1563 1550 1267 1400 1518 1537 1542 1445 1491 1652 4 1567 1428 1392 1274 1448 1448 1536 1438 1393 1584 5 1303 1244 1791 1543 1071 1350 1589 1279 1659 1491 6 1600 1488 1620 1526 1295 1028 1419 1367 1454 1447 7 1523 1317 1530 1374 1420 1305 1076 1286 1459 1305 8 1469 1414 1490 1422 1366 1407 1430 967 1535 1459 9 1624 1551 1475 1517 1556 1551 1403 1464 1077 1419 10 1638 1334 1558 1554 1461 1457 1273 1336 1386 1093 Table1 : This table shows the distances between the images of the rst series (vertical entries) and of the second series (horizontal entries). Fig.... ..."

### Table 1 The elements of G matrix in system matrix

"... In PAGE 3: ... The number of nodes of the system is 24; the number of integral units s is 2; the number of delay units z is 2; the number of branches is 37; the number of input node is 1; and the number of output node is 24. In the signal flow graph, the elements of G matrix in system matrix have the following corresponding relationship shown in Table1 , where the coefficients of the multiplier k11=-0.3090, k12=0.... ..."

### Table 1: Mapping of Elements of the Lm?1 Matrix to Elements of the Lm Matrix

"... In PAGE 8: ... All odd/even con guration pair combinations are evaluated in Table 1. Recursions of the form fi1i2 in(m) = 2 X i1;i2; ;in ai1i2 in fi1i2 in(m ? 1) can be developed for a representative spin i = i1i2 in in each equivalence class by consideration of the appropriate terms of Table1 that contribute to the summation in the right-hand-side of this equation. The f0s can be placed in the form of a linear system of di erence equations x(m) = An x(m ? 1) y(m) = TracefLn mg = Cn x(m) where the ith component of the initial state for m = 2 is given by fi1i2 in(2) = TracefPi1L2Pi2L2 PinL2Pi1g where i = i1i2 in corresponds to the appropriate representative sequence of 1 apos;s and 2 apos;s.... In PAGE 8: ... For zero eld (B = 0), the order of the system further reduces. 4 Derivation of the State Transition Matrix Table1 which relates the elements of the Lm?1 to the elements of the Lm matrix can be rewritten by de ning m ! s1, 0 m ! s2, m?1 ! s0 1, and 0 m?1 ! s0 2. The powers of the exponential term e are shown in Table 2 for B = 0.... ..."

### TABLE 1. Transition matrix elements.

1998

### Table 2. Results of the differential optimisation for the individual elements

"... In PAGE 4: ... As the purpose of the present work was to show the suitability of this method for multi-element analysis, the results of the individual optimisation are quantitatively presented and only typical examples of the results are given. The results of the differential optimisation of the target function log SNR (in dB) and the amount of information IG are shown in Table2 . The last column of the table contains the detection limit (in pg) according to the maxima of log SNR of 90 80 70 50 40 30 40 80 120 160 200 pArlhPa Fig.... ..."

### Table 2: Overloaded Distributed Matrix Operations.

"... In PAGE 4: ... It is within this le that calls to ppclient are made in order to perform the appropriate computation. Table2 lists functions... In PAGE 4: ...dense matrices. For dsparse matrices, individual elements can be retrieved and set. 3.3 p: Towards transparent constructors One problem with the constructors described in Table2 is that they do not directly corre- spond to Matlab apos;s matrix constructors such as zeros or rand. Users must learn a new set of functions for creating distributed objects.... ..."

### Table 2: Overloaded Distributed Matrix Operations.

"... In PAGE 4: ... It is within this le that calls to ppclient are made in order to perform the appropriate computation. Table2 lists functions... In PAGE 4: ...dense matrices. For dsparse matrices, individual elements can be retrieved and set. 3.3 p: Towards transparent constructors One problem with the constructors described in Table2 is that they do not directly corre- spond to Matlab apos;s matrix constructors such as zeros or rand. Users must learn a new set of functions for creating distributed objects.... ..."

### Table 2: Overloaded Distributed Matrix Operations.

"... In PAGE 4: ... It is within this le that calls to ppclient are made in order to perform the appropriate computation. Table2 lists functions... In PAGE 4: ...dense matrices. For dsparse matrices, individual elements can be retrieved and set. 3.3 p: Towards transparent constructors One problem with the constructors described in Table2 is that they do not directly corre- spond to Matlab apos;s matrix constructors such as zeros or rand. Users must learn a new set of functions for creating distributed objects.... ..."

### Table I Regression Coefficients and Standard Deviations for Expected Distances ^dij amp; Depending on k

1994