0

我正在尝试将 Alpha Beta 修剪添加到我的极小值中,但我不明白我哪里出错了。

目前我正在经历 5,000 次迭代,据一位朋友说,我应该经历大约 16,000 次迭代。选择第一个位置时,它返回 -1(失败),而此时它应该能够肯定返回 0(平局),因为它应该能够从空板中抽奖,但是我看不到当我遵循我的代码时,我哪里出错了,这似乎很好

奇怪的是,如果我在检查中切换返回 Alpha 和 Beta(以实现返回 0),计算机将尝试绘制但不会启动任何获胜动作,只会阻止

我的逻辑流程

如果我们正在寻找 alpha:如果分数 > alpha,则更改 alpha。如果 alpha 和 beta 重叠,则返回 alpha

如果我们正在寻找 beta:如果分数 < beta,则更改 beta。如果 alpha 和 beta 重叠,则返回 beta

这是我的递归调用

int MinimaxAB(TGameBoard* GameBoard, int iPlayer, bool _bFindAlpha, int _iAlpha, int _iBeta) 
{

    //How is the position like for player (their turn) on iGameBoard?
    int iWinner = CheckForWin(GameBoard);
    bool bFull = CheckForFullBoard(GameBoard);

    //If the board is full or there is a winner on this board, return the winner
    if(iWinner != NONE || bFull == true) 
    {
        //Will return 1 or -1 depending on winner
        return iWinner*iPlayer;
    }

    //Initial invalid move (just follows i in for loop)
    int iMove = -1;
    //Set the score to be instantly beaten
    int iScore = INVALID_SCORE;

    for(int i = 0; i < 9; ++i)
    {
        //Check if the move is possible
        if(GameBoard->iBoard[i] == 0) 
        {
            //Put the move in
            GameBoard->iBoard[i] = iPlayer;

            //Recall function
            int iBestPositionSoFar = -MinimaxAB(GameBoard, Switch(iPlayer), !_bFindAlpha, _iAlpha, _iBeta);

            //Replace Alpha and Beta variables if they fit the conditions - stops checking for situations that will never happen
            if (_bFindAlpha == false)
            {
                if (iBestPositionSoFar < _iBeta)
                {
                    //If the beta is larger, make the beta smaller
                    _iBeta = iBestPositionSoFar;
                    iMove = i;

                    if (_iAlpha >= _iBeta)
                    {
                        GameBoard->iBoard[i] = EMPTY;

                        //If alpha and beta are overlapping, exit the loop
                        ++g_iIterations;
                        return _iBeta;

                    }
                }
            }
            else
            {
                if (iBestPositionSoFar > _iAlpha)
                {
                    //If the alpha is smaller, make the alpha bigger
                    _iAlpha = iBestPositionSoFar;
                    iMove = i;

                    if (_iAlpha >= _iBeta)
                    {
                        GameBoard->iBoard[i] = EMPTY;

                        //If alpha and beta are overlapping, exit the loop
                        ++g_iIterations;
                        return _iAlpha;
                    }
                }
            }

            //Remove the move you just placed
            GameBoard->iBoard[i] = EMPTY;
        }
    }


    ++g_iIterations;

    if (_bFindAlpha == true)
    {
        return _iAlpha;
    }
    else
    {
        return _iBeta;
    }
}

初始调用(当计算机应该选择一个位置时)

int iMove = -1; //Invalid
int iScore = INVALID_SCORE;

for(int i = 0; i < 9; ++i) 
{
    if(GameBoard->iBoard[i] == EMPTY) 
    {
        GameBoard->iBoard[i] = CROSS;
        int tempScore = -MinimaxAB(GameBoard, NAUGHT, true, -1000000, 1000000);
        GameBoard->iBoard[i] = EMPTY;

        //Choosing best value here
        if (tempScore > iScore)
        {
            iScore = tempScore;
            iMove = i;
        }
    }
}
//returns a score based on Minimax tree at a given node.
GameBoard->iBoard[iMove] = CROSS;

任何有关我的逻辑流程的帮助将使计算机返回正确的结果并做出明智的举动将不胜感激

4

1 回答 1

1

如果没有 alpha-beta 修剪,您的算法是否可以完美运行?您的初始调用应该使用falsefor _bFindAlpha,因为根节点的行为类似于 alpha 节点,但看起来这不会产生影响:

int tempScore = -MinimaxAB(GameBoard, NAUGHT, false, -1000000, 1000000);

因此,我会建议您放弃这种_bFindAlpha废话并将您的算法转换为negamax。它的行为与 minimax 相同,但使您的代码更短更清晰。您可以在递归调用时交换和取反,而不是检查是否最大化 alpha 或最小化 beta(这与您现在可以返回函数的否定值的原因相同)。这是维基百科伪代码的略微编辑版本:

function negamax(node, α, β, player)
    if node is a terminal node
        return color * the heuristic value of node
    else
        foreach child of node
            val := -negamax(child, -β, -α, -player)
            if val ≥ β
                return val
            if val > α
                α := val
        return α

除非你喜欢单步调试搜索树,否则我认为你会发现编写一个干净、正确的 negamax 版本比调试当前的实现更容易。

于 2013-09-12T00:33:56.920 回答