baller2vec ++：一个面向协调Agent建模的前瞻性多实体Transformer
baller2vec++: A Look-Ahead Multi-Entity Transformer For Modeling Coordinated Agents
摘要：
在许多多智能体时空系统中，智能体在共享的、不可观测的变量（例如，在篮球比赛中一个队正在进行的比赛）。因此，代理的轨迹通常在任何给定的时间步是统计依赖的;然而，几乎普遍地，多代理模型隐含地假设代理的轨迹在每个时间步是统计独立的。在本文中，我们介绍了baller 2 vec ++，多实体Transformer，可以有效地协调代理建模。具体来说，baller 2 vec ++将一个专门设计的自我注意力掩模应用于位置和“前瞻”轨迹序列的混合，以学习统计相关代理轨迹的分布。我们表明，与baller 2 vec（baller 2 vec ++的前身），baller 2 vec ++可以学习模仿完美协调代理在模拟玩具数据集的行为。此外，在对职业篮球运动员的轨迹进行建模时，baller 2 vec ++的性能远远优于baller 2 vec。
英文摘要：
In many multi-agent spatiotemporal systems, agents operate under the influence of shared, unobserved variables (e.g., the play a team is executing in a game of basketball). As a result, the trajectories of the agents are often statistically dependent at any given time step; however, almost universally, multi-agent models implicitly assume the agents' trajectories are statistically independent at each time step. In this paper, we introduce baller2vec++, a multi-entity Transformer that can effectively model coordinated agents. Specifically, baller2vec++ applies a specially designed self-attention mask to a mixture of location and "look-ahead" trajectory sequences to learn the distributions of statistically dependent agent trajectories. We show that, unlike baller2vec (baller2vec++'s predecessor), baller2vec++ can learn to emulate the behavior of perfectly coordinated agents in a simulated toy dataset. Additionally, when modeling the trajectories of professional basketball players, baller2vec++ outperforms baller2vec by a wide margin.
机构、总结、中文摘要均由AI分析而来，可能存在错误，请以论文正文为准。
