### Table 1: Finite Horizon Computations

1997

"... In PAGE 6: ...Table 1: Finite Horizon Computations Table1 presents the results of the nite-horizon computations. The \stability quot; column tabulates 0 ? ( N?1 ? 1) N; which must be positive to guarantee stability (cf.... ..."

Cited by 1

### Table 3. Average RLS Heuristic Performance for Finite Planning Horizon Case * Planning Horizon (T)

"... In PAGE 16: ...1. Table3 tabulates results from the RLS heuristic. Recall that for each system there were 10 independent problem instances for each value of T, and each such problem was solved 10 times by the RLS heuristic.... In PAGE 21: ... RLS Heuristic Performance for Infinite Planning Horizon Case Table 2. Computational Requirements Table3... ..."

### Table 5. Repeated Games Results

"... In PAGE 6: ... Note that percentages do not up to 1 be- cause we have left out give- and take-type exchanges from this analysis. Table5 shows the average performance for each game. The performance of the PD agent increased from game to game, while the performance of unhelpful PD agents de- creased from game to game.... ..."

### Table 2: Coordination Game

in multiagent

"... In PAGE 6: ... If the game is iterated or repeated, the mutual cooperation total payment must exceed the temptation total payment: 2 R gt; T + S. P2 cooperates P2 defects P1 cooperates R,R S,T P1 defects T,S P,P Table 1: PD Game The pure coordination game is symmetric, two player, two strategies, with payoff matrix as given in Table2 . In the coordination game the following holds: A gt;C and D gt;B.... ..."

### Table 4 Finite horizon performance bounds P0 = Q

1997

"... In PAGE 19: ... eqn. (22)) Finally, in Table4 we calculate the nite horizon performance bounds, given by PN 2, and corresponding to P0 = Q which do not rely on our ability to calculate the optimal in nite horizon cost. Note that these bounds are not available for P0 = P 0 and P ?1 0 = 0, where PN is not monotonically non{decreasing.... In PAGE 19: ... Note that these bounds are not available for P0 = P 0 and P ?1 0 = 0, where PN is not monotonically non{decreasing. Comparing Table 3 with Table4 , we see that the nite horizon bounds given in Table 4 are quite close to those in Table 2, demonstrating the potential usefulness of this approach. 6.... In PAGE 19: ... Note that these bounds are not available for P0 = P 0 and P ?1 0 = 0, where PN is not monotonically non{decreasing. Comparing Table 3 with Table 4, we see that the nite horizon bounds given in Table4 are quite close to those in Table 2, demonstrating the potential usefulness of this approach. 6.... ..."

Cited by 3

### Table 4 Finite horizon performance bounds P0 = Q

1997

"... In PAGE 19: ... eqn. (22)) Finally, in Table4 we calculate the nite horizon performance bounds, given by PN 2, and corresponding to P0 = Q which do not rely on our ability to calculate the optimal in nite horizon cost. Note that these bounds are not available for P0 = P 0 and P ?1 0 = 0, where PN is not monotonically non{decreasing.... In PAGE 19: ... Note that these bounds are not available for P0 = P 0 and P ?1 0 = 0, where PN is not monotonically non{decreasing. Comparing Table 3 with Table4 , we see that the nite horizon bounds given in Table 4 are quite close to those in Table 2, demonstrating the potential usefulness of this approach. 6.... In PAGE 19: ... Note that these bounds are not available for P0 = P 0 and P ?1 0 = 0, where PN is not monotonically non{decreasing. Comparing Table 3 with Table 4, we see that the nite horizon bounds given in Table4 are quite close to those in Table 2, demonstrating the potential usefulness of this approach. 6.... ..."

Cited by 3

### Table 1: The perfect Bayesian equilibria of the fixed length Bayesian game. The myopic strategy for clients of both types, is to accept a price of 1 or less in both slots. The pessimistic strategy for file transferor (FT) clients is to accept no price larger than zero in the first slot, and no price larger than 2 in the second slot.

2003

Cited by 1

### Table 4. Speedup achieved for each of the three repeated substring search strategies

2003

"... In PAGE 8: ... A more detailed account of our search algorithm and the results obtained is in preparation [10]. We ran the three distributed algorithms over the aforementioned laboratory of 90 clients and recorded the speedup data shown in Table4 . For these computations we did not have sole use of the laboratory.... ..."

Cited by 5

### Table 1. Speedup achieved for each of the three repeated substring search strategies

2003

"... In PAGE 8: ... A more detailed account of our search algorithm and the results obtained is in preparation [10]. We ran the three distributed algorithms over the aforementioned laboratory of 90 clients and recorded the speedup data shown in Table1 . For these computations we did not have sole use of the laboratory.... ..."

Cited by 5

### Table 1. Speedup achieved for each of the three repeated substring search strategies.

2002

"... In PAGE 5: ... A more detailed account of these results is in preparation [10]. We ran the three distributed algorithms over the aforementioned laboratory of 90 clients and recorded the speedup data shown in Table1 . For these computations we did not have sole use of the laboratory.... ..."

Cited by 1