[link]
This paper proposes an approach to incorporate offline samples in Policy Gradient. The authors were able to do this by drawing a parallel between Policy Gradient and QLearning. This is the second paper in ICLR 2017 that tries to use offline samples to accelerate the learning in Policy Gradient. The summary of the first paper can be found [here](http://www.shortscience.org/paper?bibtexKey=journals/corr/WangBHMMKF16]). This is the first paper that describes the relationship between policy gradient and Qlearning. Lets consider a simple case of a grid world problem. In this problem, at every state, you can take one of the four actions: left, right, up, or down. Lets assume that, you are following a policy $\pi_\theta(\cdots_t)$. Let $\beta^{\pi_\theta}(s_t)$ denotes the stationary probability distribution of states under policy $\pi_\theta$ and $r(s, a)$ be the reward that we get after taking action $a$ at state $s$. Then our goal is to maximize $$ \begin{array}{ccc} &\arg\max_{\theta} \mathbb{E}_{s \sim \beta^{\pi_\theta}} \sum_{a}\pi_\theta(as) r(s, a) + \alpha H(\pi_\theta(\cdot s))&\\ & \sum_{a} \pi_\theta(as) = 1 \;\;\forall\;\; s \end{array} $$ Note that the above equation is a regularized policy optimization approach where we added a entropy bonus term, $H(\pi_{\theta}(\cdot  s))$ to enhance exploration. Mathematically, $$ H(\pi_{\theta}(\cdot  s)) =  \sum_{a} \pi_\theta(as)\log\pi_\theta(as) $$ and $$ \begin{array}{ccl} \nabla_\theta H(\pi_{\theta}(\cdot  s))& = & \sum_{a} \left(1 + \log\pi_\theta(as)\right) \nabla_\theta \pi_\theta(as) \\ & = & \mathbb{E}_{a \sim \pi_\theta(\cdots)}\left(1 + \log\pi_\theta(as)\right) \nabla_\theta \log\pi_\theta(as) \;\; \text{(likelihood trick)} \end{array} $$ Lagrange multiplier tells us that at the critical point $\theta^*$ of the above optimization problem, the gradient of optimization function should be parallel to gradient of constraint. Using the Policy Gradient Theorem, the gradient of the objective at optimal policy $\theta^*$ is $$ \mathbb{E}_{s \sim \beta^{\pi_{\theta^*}}, a \sim \pi_{\theta^*}(\cdots)} \nabla_{\theta} \log\pi_{\theta^*}(as) \left(Q^{\pi_{\theta^*}}(s, a)  \alpha  \alpha \log\pi_{\theta^*}(as)\right) $$ The gradient of the constraint at the optimal point $\theta^*$ is $$ \mathbb{E}_{s \sim \beta^{\pi_{\theta^*}}, a \sim \pi_{\theta^*}(\cdots)} \lambda_s \nabla_{\theta^*}\log\pi_{\theta^*}(as) $$ Using the theory of Lagrange multiplication $$ \begin{array}{lll} &&\mathbb{E}_{s \sim \beta^{\pi_{\theta^*}}, a \sim \pi_{\theta^*}(\cdots)} \nabla_{\theta}\log\pi_{\theta^*}(as) \left(Q^{\pi_{\theta^*}}(s, a)  \alpha  \alpha \log\pi_{\theta^*}(as)\right) = \\ &&\mathbb{E}_{s \sim \beta^{\pi_{\theta^*}}, a \sim \pi_{\theta^*}(\cdots)} \lambda_s \nabla_{\theta}\log\pi_{\theta^*}(as) \end{array} $$ If $\beta^{\pi_{\theta^*}}(s) > 0\;\; \forall\;\; s $ and $0 < \pi_{\theta^*}(a  s) < 1\;\; \forall\;\; s, a$, then for the tabular case of grid world, we get $$ Q^{\pi_{\theta^*}}(s, a)  \alpha \log\pi_{\theta^*}(as) = \lambda_{s}\;\; \forall \;\; a $$ By multiplying both sides in above equation with $\pi_{\theta^*}(as)$ and summing over $a$, we get $$ \lambda_s = V^{\pi_{\theta^*}}(s) + \alpha H(\pi_\theta(\cdot  s)) $$ Using the value of Lagrange Multiplier, the actionvalue function of the optimal policy $\pi_{{\theta^*}}$ is $$ {Q}^{\pi_{\theta^*}}(s, a) = V^{\pi_{\theta^*}}(s) + \alpha \left(H(\pi_{\theta^*}(\cdot  s)) + \log\pi_{\theta^*}(as)\right) $$ and the optimal policy $\pi_{\theta^*}$ is a softmax policy $$ \pi_{\theta^*}(as) = \exp\left( \frac{Q^{\pi_{\theta^*}}(s, a)  V^{\pi_{\theta^*}}(s)}{\alpha}  H(\pi_{\theta^*}(\cdots))\right) $$ The above relationship suggests that the optimal policy for the tabular case is a softmax policy of actionvalue function. Mainly; $$ \pi_{\theta^*}(as) = \frac{\exp\left(\frac{Q^{\pi_{\theta^*}}(s,a)}{\alpha}\right)}{\sum_b \exp\left(\frac{Q^{\pi_{\theta^*}}(s,b)}{\alpha}\right)} $$ Note that as the $\alpha \rightarrow 0$, the above policy becomes a greedy policy. Since we know that $Q$ learning reaches to an optimal policy, we can say that the above softmax policy will also converge to the optimal policy as the $\alpha \rightarrow 0$. The authors suggest that even when the policy $\pi_{\theta}$ is not an optimal policy, we can still use the $\tilde{Q}^{\pi_\theta}(s, a)$ as an estimate for actionvalue of policy $\pi_\theta$ where $$ \tilde{Q}^{\pi_{\theta}}(s, a) = V^{\pi_{\theta}}(s) + \alpha \left(H(\pi_{\theta}(\cdot  s)) + \log\pi_{\theta}(as)\right) $$ In the light of above definition of $\tilde{Q}^{\pi_\theta}(s, a)$, the update rule for the regularized policy gradient can be written as $$ \mathbb{E}_{s \sim \beta^{\pi_\theta}, a \sim \pi_\theta(\cdots)} \nabla_{\theta} \log\pi_\theta(as) \left(Q^{\pi_\theta}(s, a)  {\tilde{Q}}^{\pi_\theta}(s, a) \right) $$ Note that all the term in the above equation that depend only on the state $s$ can be incorporated using the baseline trick. **Author shows a strong similarity between Duelling DQN and Actorcritice method** In a duelling architecture, actionvalues are represented as summation of Advantage and Value function. Mathematically, $$ Q(s, a) = A^w(s, a)  \sum_a \mu(as) A^w(s, a) + V^\phi(s) $$ The goal of the middle term in the above equation to obtain advantage, $A^w(s, a)$, and value function, $V(s)$, uniquely given $Q(s, a) \forall a$. In the Duelling architecture, we will minimize the following error $$ (r+ \max_b Q(s', b)  Q(s, a))^2 $$ Consequently, we will update the $w$ and $\phi$ parameters as following: $$ \begin{array}{ccc} \Delta W &\sim& (r+ \max_b Q(s', b)  Q(s, a)) \nabla \left( A^w(s, a)  \sum_a \mu(as) A^w(s, a) \right) \\ \Delta \phi &\sim& (r+ \max_b Q(s', b)  Q(s, a)) \nabla \left( V^\phi(s) \right) \end{array} $$ Now assume an actorcritic approach, where policy is parametrized by $A^w(s, a)$ and there is a critic $V^\phi(s)$ of valuefunction. The policy $\pi_w(s, a)$ is $$ \pi_w(s, a) = \frac{e^{A^w(s, a)/\alpha}}{\sum_a e^{A^w(s, a)/\alpha}} $$ Note that $$ \nabla_w \log\pi_w(s, a) = \nabla_w \frac{1}{\alpha}\left(A^w(s, a)  \sum_a \pi_w(s, a) A^w(s, a)\right) $$ To estimate the $Q$values of policy $\pi_w$, we use the estimator obtained from optimal policy: $$ \begin{array}{ccc} \hat{Q}(s, a) &=& \alpha \left(\sum_a \pi_w(a  s)\log\pi_w(a  s) + \log\pi_w(as)\right) + V^\phi(s) \\ &=& A^w(s, a)  \sum_a \pi_w(a  s) A^w(s, a) + V^\phi(s) \end{array} $$ Note that the $Q$estimate in the actor critic and duelling architecture are different in using $\pi$ instead of $\mu$. The actor update rule in the regularized policy gradient will be $$ \Delta W \sim (r+ \max_b Q(s', b)  Q(s, a)) \nabla \left( A^w(s, a)  \sum_a \pi_w(as) A^w(s, a) \right) $$ and the critic update rule is $$ \Delta \phi \sim (r+ \max_b Q(s', b)  Q(s, a)) \nabla V^\phi(s) $$ Note that the rules to update $V^{\phi}$ is same in both DQN and actorcritic. The rule to update the critic varies in the probability distribution that is used to normalize the advantage function to make it unique. ** PGQ Algorithm** Given this information PGQ Algorithm is simple and consist of the following steps: 1. Lets $\pi_\theta(as)$ represent our policy network and $V^w(s)$ represent our value network. 2. do $N$ times 1. We collect samples $\{s_1, a_1, r_1, s_2, a_2, r_2, \cdots, s_T\}$ using policy $\pi_\theta$. 2. We compute $Q^{\pi_\theta}(s_t, a_t) = \alpha(\log \pi(a_t s_t) + H(\pi(\cdots_t))) + V(s_t)\;\; \forall t$. 3. We update $\theta$ using the regularized policy gradient approach: $$ \begin{array}{ccc} \Delta \theta & = & \nabla_\theta \log \pi_\theta(as)(r_t + \max_b \tilde{Q}^{\pi_\theta}(s_{t+1}, b)  \tilde{Q}^{\pi_\theta}(s_{t}, a_{t} )\\ \Delta \theta & = & \nabla_\theta (W_\theta(s, a)  \sum_{a} \pi_\theta(as) W_\theta(s, a))(r_t + \max_b \tilde{Q}^{\pi_\theta}(s_{t+1}, b)  \tilde{Q}^{\pi_\theta}(s_{t}, a_{t} ) \end{array} $$ where $\pi_{\theta}(s, a) = \frac{e^{W(s, a)}}{\sum_b e^{W(s, b)}}$ We update the critic by minimizing the mean square error: $$ ((r_t + \max_b \tilde{Q}^{\pi_\theta}(s_{t+1}, b)  \tilde{Q}^{\pi_\theta}(s_{t}, a_{t} ))^2 $$
Your comment:
