阿金博士采取的方法是建立起一种他称为可能行为动作的多维精确的决策空间。
His approach is to create what he calls a "multidimensional mathematical decision-space of possible behaviour actions".
学习是一种典型的强化学习，其学习效率较低，尤其是当状态空间和决策空间较大时。
Q-learning is a typical Reinforcement Learning (RL) method with a slow convergence speed especially as the scales of the state space and action space increase.
向量形式的对象描述和相似性的表示方法在某些适宜的决策空间中提供了较好的推广方法。
Vectorial object descriptions and proximity representations provide a good way for generalization in some appropriately determined Spaces.
本研究是黑龙江省攻关课题“森林资源监测与经营决策空间信息技术的研究”（GC02B608）的部分内容。
This research is part of Provincial Key Technologies R and D program of Heilongjiang: the study of the forest resources monitoring and management decision of spatial information technology (GC02B608).
本研究是黑龙江省攻关课题“森林资源监测与经营决策空间信息技术的研究”（GC02B608）的部分内容。
This research is part of Provincial Key Technologies R and D program of Heilongjiang: the study of the forest resources monitoring and management decision of spatial information technology (GC02B608).
应用推荐