In this paper, we investigate the sample complexity of policy evaluation in infinite-horizon offline reinforcement learning (also known as the off-policy evaluation problem) with linear function approximation. We identify a hard regime $d\gamma^{2}>1$, where $d$ is the dimension of the feature vector and $\gamma$ is the discount rate. In this regime, for any $q\in[\gamma^{2},1]$, we can construct a hard instance such that the smallest eigenvalue of its feature covariance matrix is $q/d$ and it requires $\Omega\left(\frac{d}{\gamma^{2}\left(q-\gamma^{2}\right)\varepsilon^{2}}\exp\left(\Theta\left(d\gamma^{2}\right)\right)\right)$ samples to approximate the value function up to an additive error $\varepsilon$. Note that the lower bound of the sample complexity is exponential in $d$. If $q=\gamma^{2}$, even infinite data cannot suffice. Under the low distribution shift assumption, we show that there is an algorithm that needs at most $O\left(\max\left\{ \frac{\left\Vert \theta^{\pi}\right\Vert _{2}^{4}}{\varepsilon^{4}}\log\frac{d}{\delta},\frac{1}{\varepsilon^{2}}\left(d+\log\frac{1}{\delta}\right)\right\} \right)$ samples ($\theta^{\pi}$ is the parameter of the policy in linear function approximation) and guarantees approximation to the value function up to an additive error of $\varepsilon$ with probability at least $1-\delta$.
翻译:在本文中, 我们用线性函数近似值来调查无限离线强化学习( 也称为离政策评估问题) 的政策评估的样本复杂性 。 我们确定一个硬制度 $d\ gamma% 2\\\ $1$, 其中美元是特性矢量的维度范围, $\ gamma$是贴现率 。 在这个制度中, 对于任何$[\ gamma% 2, 1, 我们能构建一个硬实例, 这样它特性常量变异矩阵最小的精度值是 q/ d$, 它需要$megalele( left) (\\ 离政策 left (q\\ gamma% 2\\\\\\ $ $) 美元 。 我们以离子2\\\ \\\\\\\\\\\\\\\\\\ r\\ rma\ left lex left group max lax lax lax lax lax lax lax laft sm lax lax lax lax mess lax lax lax lax lax lax lax lax max lax laxx lax laxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx = = = =xxxxxxxxxxxxxxxxxxxxxx =xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx