Newest 'policies' Questions

3

votes

0answers

35 views

What is the difference between an on-policy distribution and state visitation frequency?

On-policy distribution is defined as follows in Sutton and Barto: On the other hand, state visitation frequency is defined as follows in Trust Region Policy Optimization: $$\rho_{\pi}(s) = \sum_{t=0}^...

asked Dec 8 '21 at 10:36

user529295

3298 bronze badges

3

votes

1answer

69 views

Can an optimal policy have a value function that has a smaller value for a state than a non-optimal policy?

I'm starting to learn about the Bellman Equation and a question came to my mind. A policy $\pi$ is optimal if the value $v_\pi(s)$ is greater or equal than the value $v_{\pi'}(s)$ for all states $s \...

reinforcement-learning value-functions policies bellman-equations optimal-policy

asked Nov 8 '21 at 14:15

raphael_mav

1334 bronze badges

3

votes

1answer

64 views

In reinforcement learning, why are policies defined as functions of states and not observations?

I am new to RL and I am following Sutton & Barto's book. My doubt is, when we talk about the policy of our agent, we say it is the probability of taking some action $a$ given the state $s$. ...

reinforcement-learning definitions markov-decision-process policies

asked Sep 19 '21 at 4:44

hakiki_makato

1534 bronze badges

2

votes

1answer

52 views

Why must the value of a state under an optimal policy equal the expected return for the best action from that state?

The Sutton and Barto reinforcement learning textbook states that the value of a state under an optimal policy must equal the expected return for the best action from that state. That is, $$v_*(s) = \...

reinforcement-learning value-functions policies bellman-equations

asked Aug 22 '21 at 0:21

bonzo_pippinpaddle

376 bronze badges

1

vote

0answers

32 views

How to allow RL systems to find better policies after code changes?

Suppose that in version 1 of a reinforcement-learning system an optimal policy $A$ got generated for executing a task. But, in a newer version 2 of that application (with new code changes), there ...

reinforcement-learning policies

asked Aug 9 '21 at 10:33

user456280

1715 bronze badges

5

votes

0answers

61 views

Proof that there always exists a dominating policy in an MDP

I think that it is common knowledge that for any infinite horizon discounted MDP $(S, A, P, r, \gamma)$, there always exists a dominating policy $\pi$, i.e. a policy $\pi$ such that for all policies $\...

reinforcement-learning markov-decision-process proofs policies

asked Jul 16 '21 at 16:21

MMM

1753 bronze badges

0

votes

0answers

41 views

DQN learns to always choose the same action for all states

I have created an RL model that uses QBased policy with a neural network for estimating Q values. My action space is of 27 actions, where each action is a 3 tuple where each value can be 1, 2 or 3. ...

reinforcement-learning deep-rl dqn policies

asked May 27 '21 at 15:04

KnownUnknown

1

0

votes

1answer

63 views

Are optimal policies always deterministic, or can there also be optimal policies that are stochastic?

Let $M$ be an MDP with two states, $A$ and $B$, where $A$ is the starting state, and you always transit to the final state $B$ using two possible actions. $A_1$ gives you rewards that are normally ...

markov-decision-process policies optimality stochasticity

asked May 1 '21 at 5:19

Abc1729

353 bronze badges

1

vote

1answer

54 views

Is the policy gradient expression in Fundamentals of Deep Learning wrong?

I don't understand the policy gradient as explained in Chapter-9 (Deep Reinforcement Learning) of the book Fundamentals of deep learning. Here is the whole paragraph: Policy Learning via Policy ...

deep-rl objective-functions math policy-gradients policies

asked Apr 8 '21 at 8:04

Jonathan Sands

113 bronze badges

3

votes

1answer

186 views

How can we find the value function by solving a system of linear equations without knowing the policy?

An MDP is a Markov Reward Process with decisions, it’s an environment in which all states are Markov. This is what we want to solve. An MDP is a tuple $(S, A, P, R, \gamma)$, where $S$ is our state ...

markov-decision-process policies bellman-equations transition-model

asked Feb 16 '21 at 8:51

Abc1729

353 bronze badges

2

votes

1answer

72 views

Why do I get the best policy before Q values converge using DQN?

I have implemented DQN algorithm and wonder why during testing, the best performance is achieved by a policy from about 300 episode, when mean Q values converge at about 800 episode? Mean Q-values ...

dqn deep-rl convergence value-functions policies

asked Feb 4 '21 at 20:48

opal

314 bronze badges

2

votes

1answer

52 views

Is a learned policy, for a deterministic problem, trained in a supervised process, a stochastic policy?

If I trained a neural network with 4 outputs (one for each action: move down, up, left, and right) to move an agent through a grid (deterministic problem). The output of the neural network is a ...

neural-networks policies deterministic-policy stochastic-policy softmax-policy

asked Feb 3 '21 at 12:47

Xtalker

192 bronze badges

2

votes

1answer

256 views

What is a "learned policy" in Q-learning?

I am completing an assignment at the moment. One of the assignment questions asks how you identified the learned policy and how you obtained it. The question is a reinforcement learning question, and ...

reinforcement-learning q-learning terminology value-functions policies

asked Jan 25 '21 at 9:32

user43972

2

votes

0answers

29 views

Does there necessarily exist "dominated actions" in a MDP?

In a Markov Decision Process, is it possible that there exists no "dominated action"? I define a dominated action the following way: we say that $(s,a)$ is a dominated action, if $\forall \...

reinforcement-learning markov-decision-process proofs value-functions policies

asked Jan 9 '21 at 19:27

Phil

313 bronze badges

1

vote

0answers

15 views

How to obtain the policy in the form of a finite-state controller from the value function vectors over the belief space of the POMDP?

I was reading this paper by Hansen. It says the following: A correspondence between vectors and one-step policy choices plays an important role in this interpretation of a policy. Each vector in $\...

papers value-functions policies pomdp policy-iteration

asked Dec 25 '20 at 8:41

Rnj

2111 silver badge6 bronze badges

1

vote

0answers

80 views

What kind of reinforcement learning method does AlphaGo Deepmind use to beat the best human Go player?

In reinforcement learning, there are model-based versus model-free methods. Within model-based ones, there are policy-based and value-based methods. AlphaGo Deepmind RL model has beaten the best Go ...

reinforcement-learning policies model-based-methods model-free-methods value-based-methods

asked Dec 23 '20 at 6:14

user781486

1835 bronze badges

3

votes

1answer

157 views

Are there some notions of distance between two policies?

I want to determine some distance between two policies $\pi_1 (a \mid s)$ and $\pi_2 (a \mid s)$, i.e. something like $\vert \vert \pi_1 (a \mid s) - \pi_2(a \mid s) \vert \vert$, where $\pi_i (a\mid ...

reinforcement-learning reference-request policies kl-divergence wasserstein-metric

asked Dec 13 '20 at 16:20

Felix P.

2871 silver badge6 bronze badges

0

votes

1answer

67 views

How to use and update a shared/global policy between Reinforcement Learning Agents

I would be grateful for some guidance on a RL problem I am trying to solve where multiple RL agents use a common/global policy at the initial state of an episode in the RL Environment, and then update ...

reinforcement-learning environment policies

asked Nov 25 '20 at 12:14

RL_NOOB

153 bronze badges

5

votes

2answers

348 views

Given two optimal policies, is an affine combination of them also optimal?

If there are two different optimal policies $\pi_1, \pi_2$ in a reinforcement learning task, will the linear combination (or affine combination) of the two policies $\alpha \pi_1 + \beta \pi_2, \alpha ...

reinforcement-learning proofs policies optimal-policy optimality

asked Nov 18 '20 at 7:04

yang liu

533 bronze badges

0

votes

0answers

48 views

Off-policy full-random training in easy-to-explore environment

Let say we are in an environment where a random agent can easily explore all the states of an environment (for example: tic-tac-toe). In those environments, using off-policy algorithm, is it a good ...

reinforcement-learning dqn policies off-policy-methods epsilon-greedy-policy

asked Nov 7 '20 at 17:59

Loheek

2661 silver badge6 bronze badges

3

votes

1answer

149 views

Why do we have two similar action selection strategies for UCB1?

In the literature, there are at least two action selection strategies associated with the UCB1's action selection strategy/policy. For example, in the paper Algorithms for the multi-armed bandit ...

reinforcement-learning policies sutton-barto multi-armed-bandits upper-confidence-bound

asked Oct 23 '20 at 10:32

nbro♦

31.7k8 gold badges66 silver badges131 bronze badges

1

vote

0answers

38 views

Dynamically adapting activation function

I am training a network through reinforcement learning. The policy network learns rotations, but depending on the actual input (state), the output of the network should be restricted to be in certain ...

neural-networks reinforcement-learning activation-functions policies state-of-the-art

asked Oct 19 '20 at 22:12

Ardi

312 bronze badges

2

votes

0answers

35 views

What's an example of a simple policy but a complex value function?

Hado van Hasselt, a researcher at DeepMind, mentioned in one of his videos (from 7:20 to 8:20) on Youtube (about policy gradient methods) that there are cases when the policy is very simple compared ...

reinforcement-learning policy-gradients policies value-functions applications

asked Aug 27 '20 at 10:16

connected-subgroup

9311 silver badge13 bronze badges

2

votes

1answer

61 views

Why doesn't value iteration use $\pi(a \mid s)$ while policy evaluation does?

I was looking at the Bellman equation, and I noticed a difference between the equations used in policy evaluation and value iteration. In policy evaluation, there was the presence of $\pi(a \mid s)$, ...

reinforcement-learning policies value-iteration policy-iteration bellman-equations

asked Aug 25 '20 at 12:35

Chukwudi Ogbonna

954 bronze badges

2

votes

0answers

26 views

Do we assume the policy to be deterministic when proving the optimality?

In reinforcement learning, when we talk about the principle of optimality, do we assume the policy to be deterministic?

reinforcement-learning proofs policies deterministic-policy

asked Aug 18 '20 at 9:32

hakiki_makato

1534 bronze badges

1

vote

2answers

142 views

Why is the optimal policy for an infinite horizon MDP deterministic?

Could someone please help me gain some intuition as to why the optimal policy for a Markov Decision Process in the infinite horizon case (agent acts forever) is deterministic?

reinforcement-learning markov-decision-process policies optimal-policy optimality

asked Aug 6 '20 at 7:42

connected-subgroup

9311 silver badge13 bronze badges

2

votes

1answer

70 views

Is a policy in reinforcement learning analogous to a field such as APF?

If a policy maps states to actions in reinforcement learning, then for a path planning with obstacles, can't we simply use Artificial Potential Field fields for path planning and model policy ...

reinforcement-learning markov-decision-process policies path-planning

asked Jul 28 '20 at 15:19

gfdsal

1716 bronze badges

2

votes

1answer

72 views

Is it common to have extreme policy's probabilities?

I have implemented several policy gradient algorithms (REINFORCE, A2C, and PPO) and am finding that the resultant policy's action probability distributions can be rather extreme. As a note, I have ...

reinforcement-learning policy-gradients policies

asked Jul 20 '20 at 21:11

curiouscat22

333 bronze badges

3

votes

1answer

128 views

What is meant by "generate the data" in describing the difference between on-policy and off-policy?

From the book: Sutton, Richard S.,Barto, Andrew G.. Reinforcement Learning (Adaptive Computation and Machine Learning series) (p. 100). The MIT Press. Kindle Edition. " following is stated: "...

reinforcement-learning policies off-policy-methods

asked Jul 12 '20 at 14:15

blue-sky

2971 silver badge9 bronze badges

1

vote

1answer

89 views

Do we need multiple parallel environments to train in batches an on-policy algorithm?

When using an on-policy method in reinforcement learning, like advantage actor-critic, you shouldn't use old data from an experience buffer, since a new policy requires new data. Does this mean that ...

reinforcement-learning environment policies experience-replay on-policy-methods

asked Jul 2 '20 at 19:46

Daniel

1112 bronze badges

1

vote

0answers

81 views

Does the off-policy evaluation work for non-stationary policies?

As the title says, in reinforcement learning, does the off-policy evaluation work for non-stationary policies? For example, IS (importance sampling)-based estimators, such as weighted IS or doubly ...

reinforcement-learning policies off-policy-methods importance-sampling

asked Jun 21 '20 at 4:21

Hunnam

2071 silver badge5 bronze badges

2

votes

2answers

226 views

Why is the policy not a part of the MDP definition?

I'm reading an article on reinforcement learning, and I don't understand why the agent's policy $\pi$ is not part of definition of Markov Decision process(MDP): Bu, Lucian, Robert Babu, and Bart De ...

reinforcement-learning definitions markov-decision-process policies

asked May 23 '20 at 19:03

Raphael Augusto

1877 bronze badges

3

votes

1answer

373 views

Why do RL implementations converge on one action?

I have seen this happening in implementations of state-of-the-art RL algorithms where the model converges to a single action over time after multiple training iterations. Are there some general ...

reinforcement-learning convergence policies

asked May 10 '20 at 15:39

JAYDEEP GHOSE

511 bronze badge

3

votes

2answers

121 views

How does the AlphaGo Zero policy decide what move to execute?

I was going through the AlphaGo Zero paper and I was trying to understand everything, but I just can't figure out this one formula: $$ \pi(a \mid s_0) = \frac{N(s_0, a)^{\frac{1}{\tau}}}{\sum_b N(s_0,...

reinforcement-learning policies deepmind alphago-zero alphago

asked May 7 '20 at 11:09

Eloi M.

334 bronze badges

2

votes

3answers

79 views

How to estimate a behavior policy for off-policy learning based on data?

I have a dataset which includes states, actions, and reward. The dataset includes information on the transition, i.e., $p(r,s' \mid s,a)$. Is there a way to estimate a behavior policy from this ...

reinforcement-learning policies off-policy-methods

asked Apr 27 '20 at 14:44

Tracy Yang

2812 silver badges4 bronze badges

3

votes

2answers

830 views

Why do value iteration and policy iteration obtain similar policies even though they have different value functions?

I am trying to implement value and policy iteration algorithms. My value function from policy iteration looks vastly different from the values from value iteration, but the policy obtained from both ...

reinforcement-learning comparison policies value-iteration policy-iteration

asked Apr 21 '20 at 20:03

PyWalker27

2151 gold badge2 silver badges6 bronze badges

2

votes

1answer

83 views

Why do I need an initial arbitrary policy to implement value iteration algorithm

I've been recently given an assignment based on Reinforcement Learning and I'm supposed to implement the value iteration algorithm in a grid environment. The assignment: My doubt is why do I even ...

reinforcement-learning policies value-functions value-iteration

asked Apr 16 '20 at 12:11

Carrick

1215 bronze badges

2

votes

2answers

329 views

Is my understanding of the value function, Q function, policy, reward and return correct?

I'm a beginner in the RL field, and I would like to check that my understanding of certain RL concepts. Value function: How good it is to be in a state S following policy π. ...

reinforcement-learning q-learning terminology policies return

asked Apr 16 '20 at 1:53

BG10

1156 bronze badges

1

vote

2answers

78 views

Why does the policy $\pi$ affect the Q value?

From my understanding, the policy $\pi$ is basically how the agent acts (i.e. the actions it will take in each state). However, I am confused about the Q value and how it is "affected" by a policy. ...

reinforcement-learning q-learning policies value-functions

asked Apr 16 '20 at 0:52

BG10

1156 bronze badges

3

votes

0answers

57 views

Representation of state space, action space and reward system for Reinforcement Learning problem

I am trying to solve the problem of an agent dynamically discovering(start with no information about the environment) the environment and to explore as much of the environment as possible without ...

reinforcement-learning dqn policy-gradients policies

asked Mar 22 '20 at 20:22

Incompleteness

312 bronze badges

0

votes

0answers

80 views

How to understand and visualize a trained RL agent's policy when the state space is high dimensional?

What are typical ways to understand and visualize a trained RL agent's policy when the state space is of high dimension (but not images)? For example, suppose state and action are denoted by $s=(...

reinforcement-learning markov-decision-process policies

asked Mar 3 '20 at 16:39

Ryan

1214 bronze badges

3

votes

3answers

348 views

Is the policy really invariant under affine transformations of the reward function?

In the context of a Markov decision process, this paper says it is well-known that the optimal policy is invariant to positive affine transformation of the reward function On the other hand, ...

reinforcement-learning markov-decision-process policies reward-functions reward-shaping

asked Mar 3 '20 at 0:18

riceissa

1713 bronze badges

2

votes

0answers

50 views

Can the importance sampling estimator have a non-stationary behaviour policy even if the target policy is stationary?

The inverse propensity score (IPS) estimator, which is used for off-policy evaluation in a contextual bandit problem, is well explained in the paper Doubly Robust Policy Evaluation and Optimization. ...

reinforcement-learning comparison policies off-policy-methods importance-sampling

asked Jan 21 '20 at 19:53

Hunnam

2071 silver badge5 bronze badges

2

votes

0answers

93 views

AlphaGo Zero: Does the policy head give a probability for every possible move?

If I understood correctly, the AlphaGo Zero network returns two values: a vector of logit probabilities p and a value v. My question is: in this vector that it is outputted, do we have a probability ...

deep-learning policies deepmind alphago-zero

asked Nov 11 '19 at 1:33

ihavenoidea

2451 silver badge8 bronze badges

4

votes

1answer

410 views

Why does having a fixed policy change a Markov Decision Process to a Markov Reward Process?

If a policy is fixed, it is said that a Markov Decision Process (MDP) becomes a Markov Reward Process (MRP). Why is this so? Aren't the transitions and rewards still parameterized by the action and ...

reinforcement-learning markov-decision-process policies markov-reward-process

asked Sep 30 '19 at 2:15

Peter

433 bronze badges

4

votes

2answers

136 views

Why is having low variance important in offline policy evaluation of reinforcement learning?

Intuitively, I understand that having an unbiased estimate of a policy is important because being biased just means that our estimate is distant from the truth value. However, I don't understand ...

reinforcement-learning policies bias-variance-tradeoff

asked Sep 29 '19 at 15:36

Hunnam

2071 silver badge5 bronze badges

4

votes

1answer

236 views

Can someone please help me validate my MDP?

Problem Statement : I have a system with four states - S1 through S4 where S1 is the beginning state and S4 is the end/terminal state. The next state is always better than the previous state i.e if ...

reinforcement-learning rewards policies markov-decision-process

asked Sep 5 '19 at 15:08

Bhavana

836 bronze badges

1

vote

0answers

62 views

Finding optimal Value function and Policy for an MDP

I am solving an RL MDP problem which is model based. I have an MDP which has four possible states S1-S4 and four different actions A1-A4, with S4 being terminal state and S1 is the beginning state. ...

reinforcement-learning rewards policies markov-decision-process

asked Sep 4 '19 at 10:52

Bhavana

836 bronze badges

1

vote

1answer

40 views

Do I need to store the policy for RL?

I am creating a zero-sum game with RL and wondered if I need to store the policy, or if there are other RL methods that produce similar results (consistently beating the human player) without the need ...

reinforcement-learning policies off-policy-methods storage

asked Aug 10 '19 at 16:06

mason7663

5732 silver badges7 bronze badges

2

votes

1answer

51 views

If deep Q learning involves adjusting the value function for a specific policy, then how do I choose the right policy?

I wrote a simple implementation of Flappy Bird in Python, and now I'm trying to train an agent to play it at a reasonable skill level using TFLearn. I feed the network an input vector of size 4: ...

q-learning deep-rl policies

asked Jul 13 '19 at 21:48

sOvr9000

336 bronze badges

Questions tagged [policies]

Related Tags