问题描述
我的 Minimax 算法运行良好,我什至通过记录值的层次结构来检查它。但是,当我尝试评估板时(当我最初作为函数参数提供的深度被命中时),我这样做的方式效率低下。我也试过另一种方式(我会把它放在本文下面),但这一次,它是错误的。
这是评估的初始代码(我认为它是原始的(有点)因为它确实评估了可以采取的潜在行动,而不是整个董事会):
// Here's the board with every square evaluated
const sq_val = [
[160,-20,20,5,160],[-20,-40,-5,-20],[20,15,3,20],[5,5],[160,160]
]
let elementVals = (AI === "black") ? 0 : 1
if (!isMaximizing) {
elementVals = (AI === "black") ? 1 : 0
}
const movesAvailable = checkSquaresForSides(my_game_board)[elementVals] //All the potential moves for the current player (AI if it is the maximizing player's turn,our player if not)
let bestscore = 0 // The value that we'll return
// In this code,it checks the value of every potential move and set the value of bestscore to the highest one
for (var moveMax=0;moveMax<movesAvailable.length;moveMax++) {
const coord = movesAvailable[moveMax].coordinates
if (coord) {
const value = sq_val[coord[0]][coord[1]]
bestscore += value
}
}
// If it is the minimizing player's turn,since it's best move would be the worst case for us,get the opposite of the value
bestscore = (isMaximizing) ? bestscore : -bestscore
// Return the value
return bestscore
这是我尝试的另一种方法:
// declare the maximizing evaluation board
const sq_val = [
[100,100],[100,100]
]
// Get all the stones on the board (black and white seperated)
const allStonesSep = []
for (var row=0;row<8;row++) {
for (var col=0;col<8;coL++) {
let elem = board[row][col]
if (isMaximizing) {
if (elem === AI) {allStonesSep.push(sq_val[row][col])}
} else {
if (elem !== ourPlayer) {allStonesSep.push(sq_val[row][col])}
}
}
}
// declare the bestscore
let bestscore = allStonesSep.reduce((a,b) => a + b,0)
if (!isMaximizing) {
bestscore = -bestscore
}
// Handle the value depending on the maximizing player color and the value of maximizing
// Return the value
return bestscore
我觉得没必要放极小极大算法代码;但如果您认为有必要,请发表评论。
解决方法
暂无找到可以解决该程序问题的有效方法,小编努力寻找整理中!
如果你已经找到好的解决方法,欢迎将解决方案带上本链接一起发送给小编。
小编邮箱:dio#foxmail.com (将#修改为@)