aditijc/snooker-testbed-phase4f-discrete-v1

该数据集是斯诺克测试平台第4F阶段的突破性运行结果，通过将连续的动作空间离散化为7200个动作，成功突破了之前13次运行中98%犯规率的瓶颈。数据集包含10行8列，记录了PPO时间步、课程阶段、平均得分、最大得分、平均击球数、平均效率、平均犯规率和回合数等关键指标。实验结果表明，离散化动作空间显著提高了性能，平均得分提高了2倍，犯规率首次降至97%以下。

This dataset presents the breakthrough experimental results from the 4F stage of the snooker test platform. By discretizing the continuous action space into 7200 actions, the bottleneck of 98% foul rate encountered in the prior 13 trial runs has been successfully overcome. The dataset includes 10 rows and 8 columns, documenting critical metrics including PPO time steps, course stage, average score, maximum score, average number of shots, average efficiency, average foul rate, and number of rounds. Experimental findings demonstrate that discretizing the action space substantially enhances performance: the average score has doubled, and the foul rate has dropped below 97% for the first time.