[其他] ICML-Efficient Deviation Types and Learning for Hindsight Rationality in Ex ... [推广有奖]

1关注
168
粉丝

已卖：5761份资源

泰斗

还不是VIP/贵宾

威望: 0 级
论坛币: 13312 个
通用积分: 770.7901
学术水平: 460 点
热心指数: 485 点
信用等级: 415 点
经验: 117019 点
帖子: 4686
精华: 0
在线时间: 16742 小时
注册时间: 2013-1-2
最后登录: 2026-4-18

楼主

Mujahida

发表于 2025-8-10 15:04:56 |AI写论文

是否 +2 论坛币

k人参与回答

经管之家送您一份

应届毕业生专属福利!

求职就业群

赵安豆老师微信：zhaoandou666

经管之家联合CDA

送您一个全额奖学金名额~ !

立即领取

感谢您参与论坛问题回答

经管之家送您两个论坛币！

+2 论坛币

Effcient Deviation Types and Learning for Hindsight Rationality in
            Extensive-Form Games

Dustin Morrill 1 Ryan D’Orazio 2 Marc Lanctot 3 James R. Wright 1 Michael Bowling 1 3 Amy R. Greenwald 4

            Abstract                measured by regret in hindsight relative to strategy transfor-
                                 mations, also called deviations, rather than as prospective
Hindsight rationality is an approach to playing       optimality with respect to belief ...