Tools
Sorted by:
Try your query at:
 |
 |
 |
 |
 |
 |
Results 1 - 10
of
10,258
Table 2. TextGame Analysis results
2005
Cited by 3
Table 1. Fitness points awarded by the aggregate success/failure mode Fmode_2, for pairs of reciprocal games during a generational tournament.
Table 1. Fitness points awarded by the aggregate success/failure mode Fmode_2, for pairs of reciprocal games during a generational tournament.
Table 1. Fitness points awarded by the aggregate success/failure mode Fmode_2, for pairs of reciprocal games during a generational tournament.
Table 2 illustrates the lexicon of case-base CBj from Figure 1 during the learning phase. In the table, u refers to the number of times the label has been used in different language games by this agent while, a refers to how often it has been successfully used. We can see that the relationship between label l1 and case cj9 has been successfully used 8 times in 10 different language games while the relationship between l1 and case cj6 has been successful only once in 8 games.
2005
"... In PAGE 8: ... Table2 . The lexicon of case base CBj during the learning phase.... ..."
Cited by 2
Table I This theory is very efficient in simple games such as the game of Nim. It has been applied with success to the endgame of Go [Wolfe 1991]. But this theory requires the subgames of a same position to be independent, it also requires to be able to forecast all the possible moves until the end of the game. These assumptions do not apply in complex games such as Go and Chess. I show in this article that the introduction and the management of uncertainty in Combinatorial Game Theory permits to use it in complex games.
1996
Cited by 3
Table I This theory is very efficient in simple games such as the game of Nim. It has been applied with success to the endgame of Go [Wolfe 1991]. But this theory requires the subgames of a same position to be independent, it also requires to be able to forecast all the possible moves until the end of the game. These assumptions do not apply in complex games such as Go and Chess. I show in this article that the introduction and the management of uncertainty in Combinatorial Game Theory permits to use it in complex games.
1996
Cited by 3
Table 1. Classification of the video games
"... In PAGE 5: ...roups. In each group, operations are directly mapped into commands. By shift- ing these groups, objective commands can be achieved, with a smaller number of operations. In this paper, games are classified according to response speed of operation and the numbers of commands (see Table1 ). In Table 1, games with a large number of commands and quick response speed are classified as Type 1.... In PAGE 7: ...), and the Othello game (SUCCESS Corp.) was used for ex- periments, which can be classified into Type 4 (see Table1 ). In this setup, five commands are necessary for operation: up, down, left, right, and stone arrangement.... ..."
Table 2. The result of 2500 trials in 15 full games .In the leftmost column the estimation of success rate of the passes using the constructed rules by Ant-Miner algorithm are given whereas the rightmost column shows the real success rate of the passes in percent
"... In PAGE 5: ....8-0.9 0.7-0.8 0.6-0.7 0.8-0.9 0.7-0.8 0.6-0.7 Number 1050 3485 185 1318 2913 391 Success% Failure% Miss% 79 15 5 63 29 8 58 31 10 82 12 6 69 23 8 64 33 3 Also, we investigated the results of the passes in a full game situation within 2500 passes using constructed rules. The results are given in Table2 . Since just the successful passes are ideal in a real game, we recorded just the number of successful passes and their associated estimation of success rate.... ..."
Table 5 shows the average performance for each game. As ex- pected, the performance of the MP agent increased from game to game, while the performance of unhelpful PD agents decreased from game to game. This result is supported by the fact that the MP agent avoided interacting with unhelpful agents, as shown in Table 4. Unhelpful PDs do worse from game to game, and are always worse off then helpful PDs, indicating that the MP agent successfully adapted to these agents. Although the performance of the MP agent was consistently better that the performance of all agents in every game, we were surprised that the performance of both helpful and unhelpful SP agents increased from game 2 to game 3. We hypothesized that some of the SP agents were exploit- ing the MP agent and as a result, increasing their average score.
"... In PAGE 5: ...96 91.46 Table5 : Agent performance by game In contrast, helpful SP agents were more likely to engage in reciprocal exchanges (65%) and far less likely to remain idle to- wards others (1%), regardless of their class, indicating that they were more vulnerable to exploitation. Note that percentages do not add up to 100% because we have left out exchanges which were not reciprocal.... ..."
Results 1 - 10
of
10,258
