Questions tagged [reinforcement-learning]

For questions related to learning controlled by external positive reinforcement or negative feedback signal or both, where learning and use of what has been thus far learned occur concurrently.

Filter by
Sorted by
Tagged with
0
votes
0answers
27 views

Why in RL function approximators with recurrent structures can learn planning?

In the paper An Investigation of Model-Free Planning the authors use ConvLSTM to learn a planning function. In particular, for each input x_t at time-step ...
2
votes
1answer
86 views

Non-differentiable reward function to update a neural network

In Reinforcement Learning, when reward function is not differentiable, a policy gradient algorithm is used to update the weights of a network. In the paper Neural Architecture Search with ...
0
votes
0answers
67 views

Should the network weights converge when training Deep Q networks?

I have two sets of data, one training and one test set. I use the train set to train the deep q network model variant. I also continuously evaluate the agent Q values obtained on the test set every ...
5
votes
1answer
169 views

How do I recognise a bandit problem?

I'm having difficulty understanding the distinction between a bandit problem and a non-bandit problem. An example of the bandit problem is an agent playing $n$ slot machines with the goal of ...
3
votes
1answer
122 views

How do we express $q_\pi(s,a)$ as a function of $p(s',r|s,a)$ and $v_\pi(s)$?

The task (exercise 3.13 in the RL book by Sutton and Barto) is to express $q_\pi(s,a)$ as a function of $p(s',r|s,a)$ and $v_\pi(s)$. $q_\pi(s,a)$ is the action-value function, that states how good ...
2
votes
0answers
29 views

If the minimum Q value is decreasing and the maximum Q value increasing, is this a sign that dueling double DQN is diverging?

I'm training a dueling double DQN agent with prioritized replay buffer and notice that the min Q values are decreasing, while the max Q values are increasing. Is this a sign that it is diverging? ...
2
votes
1answer
63 views

How does $\mathbb{E}$ suddenly change to $\mathbb{E}_{\pi'}$ in this equation?

In Sutton-Barto's book on page 63 (81 of the pdf): $$\mathbb{E}[R_{t+1} + \gamma v_\pi(S_{t+1}) \mid S_t=s,A_t=\pi'(s)] = \mathbb{E}_{\pi'}[R_{t+1} + \gamma v_\pi(S_{t+1}) \mid S_{t} = s]$$ How does $...
1
vote
2answers
61 views

Should I use exploration strategy in Policy Gradient algorithms?

In policy gradient algorithms the output is a stochastic policy - a probability for each action. I believe that if I follow the policy (sample an action from the policy) I make use of exploration ...
1
vote
1answer
81 views

Can I apply DQN or policy gradient algorithms in the contextual bandit setting?

I have a problem which I believe can be described as a contextual bandit. More specifically, in each round, I observe a context from the environment consisting of five continuous features, and, ...
2
votes
2answers
82 views

How to convert sequences of images into state in DQN?

I recently read the DQN paper titled: Playing Atari with Deep Reinforcement Learning. My basic and rough understanding of the paper is as follows: You have two neural networks; one stays frozen for a ...
5
votes
2answers
71 views

Why state-action value function as an expected value of the return and state value function, does not need to follow policy?

I often see, the state-action value function is expressed as: $q_{\pi}(s,a)=\mathbb{E}_{\pi}[R_{t+1}+\gamma G_{t+1} | S_t=s, A_t = a] = \mathbb{E}[R_{t+1}+\gamma v_{\pi}(s') |S_t = s, A_t =a]$ Why ...
1
vote
1answer
73 views

How to implement RAM versions of Atari games

I have coded the breakout RAM version, but, unfortunately, its highest reward was 5. I trained it for about 2 hours and never reached a higher score. The code is huge, so I can't paste here, but, in ...
2
votes
1answer
56 views

Why does AlphaGo Zero select move based on exponentiated visit count?

From the AlphaGo Zero paper, AlphaGo Zero uses an exponentiated visit count from the tree search. Why use visit count instead of the mean action value $Q(s, a)$?
3
votes
0answers
68 views

How to take actions at each episode and within each step of the episode in deep Q learning?

In deep Q learning, we execute the algorithm for each episode, and for each step within an episode, we take an action and record a reward. I have a situation where my action is 2-tuple $a=(a_1,a_2)$. ...
2
votes
1answer
120 views

How to train a reinforcement learning agent from raw pixels?

How would you train a reinforcement learning agent from raw pixels? For example, if you have 3 stacked images to sense motion, then how would you pass them to neural networks to output Q-learning ...
3
votes
1answer
38 views

What's the right way of building a deep Q-network?

I'm new to RL and to deep q-learning and I have a simple question about the architecture of the neural network to use in an environment with a continous state space a discrete action space. I tought ...
3
votes
1answer
62 views

What is the solution to the exercise 3.11 in the RL book by Sutton and Barto

I'm trying to solve the exercise 3.11 from the second-edition from the Sutton and Barto's book Exercise 3.11 If the current state is $S_t$ , and actions are selected according to a stochastic ...
0
votes
0answers
25 views

How to choose hyperparameters in double DQN?

I'm looking for some indications about the tuning of hyperparameters in building my double DQN. I have a time series problem (with about 2000 observations and no terminal state, I have to max the ...
1
vote
1answer
32 views

When discounted MAB is useful?

Many of multi-armed bandit algorithms are used when the total reward is the sum of all rewards. However, in RL, the discounted reward is mainly used. Why the discounted reward is not prevailing in MAB ...
4
votes
2answers
88 views

Why is $G_{t+1}$ is replaced with $v_*(S_{t+1})$ in the Bellman optimality equation?

In equation 3.17 of Sutton and Barto's book: $$q_*(s, a)=\mathbb{E}[R_{t+1} + \gamma v_*(S_{t+1}) \mid S_t = s, A_t = a]$$ $G_{t+1}$ here have been replaced with $v_*(S_{t+1})$, but no reason has ...
3
votes
2answers
55 views

Is there any difference between reward and return in reinforcement learning?

I am reading Sutton and Barto's book on reinforcement learning. I thought that reward and return were the same things. However, in Section 5.6 of the book, 3rd line, first paragraph, it is written: ...
2
votes
1answer
55 views

How is $\Delta$ updated in true online TD($\lambda$)?

In the RL textbook by Sutton & Barto section 7.4, the author talked about the "True online TD($\lambda$)". The figure (7.10 in the book) below shows the algorithm. At the end of each step, $V_{...
1
vote
2answers
54 views

Are there any good tutorials about training RL agent from raw pixels using PyTorch?

Is there any good tutorials about training reinforcement learning agent from raw pixels using PyTorch? I don't understand the official PyTorch tutorial. I want to train the agent on the atari ...
0
votes
0answers
15 views

Is there any advantage to using a non-diagonal covariance matrix for a policy distribution?

For reinforcement learning implementations with a gym.spaces.Box action space, which is the product of $k$ real closed intervals, it is common (actually more like ...
0
votes
1answer
38 views

What is correct update when the some indexes are not available?

To update the Q table Q-learning takes the arg max of the Q values - the state, value mappings. For example, in tic tac toe the state XOX OXO -X- contains two ...
1
vote
0answers
62 views

How can the target rely on untrained parameters?

I'm trying to understand DQN. I understand where the loss function comes from. I'm just unsure about why the target function works in practice. Given the loss function $$ L_i(\theta_i) = [(y_i - Q(s,a;...
2
votes
1answer
50 views

What is the difference between reinforcement learning and evolutionary algorithms?

What is the difference between reinforcement learning (RL) and evolutionary algorithms (EA)? For some problems, you could presumably co-evolve two "species" populations using evolutionary algorithms ...
2
votes
0answers
24 views

Are the final states not being updated in this $n$-step Q-Learning algorithm?

I am reading this paper and in algorithm 3 they describe an $n$-step Q-Learning algorithm. Below is the pseudo-code. From this pseudo-code, it looks as though the final tuples that they would ...
1
vote
1answer
35 views

What is meant by the rank of the scoring function here?

I've been reading this paper on Knowledge Graph Reasoning for Explainable Recommendation lately, and I don't understand a particular section: Specifically, the scoring function $f((r,e)|u)$ maps any ...
1
vote
1answer
55 views

Is there any programming practice website for beginners in Reinforcement Learning [closed]

I am doing an online course on Reinforcement Learning from university of Alberta. It focus too much on theory. I am engineering and I am interested towards applying RL to my applications directly. ...
1
vote
0answers
39 views

Calculating the advantage 'gain' of actions in model-free reinforcement learning

I have a simple question about model-free reinforcement. In a model I'm writing about, I want to know the value 'gain' we'd get for executing an action, relative to the current state. That is, what ...
1
vote
0answers
15 views

How to learn how to select a subgraph via reinforcement learning?

I have the following problem. I am given a graph with a lot (>30000) nodes. Nodes are associated with a low (<10)-dimensional feature vector, and edges are associated with a low (<10)-...
2
votes
1answer
32 views

Can the agent wait until the end of the episode to determine the reward in SARSA?

From Sutton and Barto's book Reinforcement Learning (Adaptive Computation and Machine Learning series) (p. 99), the following definition for first-visit MC prediction, for estimating $V \sim V_\pi$ is ...
1
vote
1answer
78 views

What is a RAM state in the gym's breakout-ram environment?

I have encountered the gym environment and decided to create AI that plays breakout. Here is the link: https://gym.openai.com/envs/Breakout-ram-v0/. The documentation says that the state is ...
1
vote
1answer
74 views

Can tabular Q-learning converge even if it doesn't explore all state-action pairs?

My understanding of tabular Q-learning is that it essentially builds a dictionary of state-action pairs, so as to maximize the Markovian (i.e., step-wise, history-agnostic?) reward. This incremental ...
1
vote
0answers
29 views

Actor-Critic implementation not learning

I've implemented a vanilla actor-critic and have run into a wall. My model does not seem to be learning the optimal policy. The red graph below shows its performance in cartpole, where the algorithm ...
1
vote
0answers
35 views

How to prevent deep Q-learning algorithms to overfit?

I have recently solved the Cartpole problem using double deep Q-learning. When I saw how the agent was doing, it used to go right every time, never left, and it did similar actions all the time. Did ...
1
vote
1answer
36 views

If deep Q-learning starts to choose only one action, is this a sign that the algorithm diverged?

I'm working on a deep q-learning model in an infinite horizon problem, with a continous state space and 3 possible actions. I'm using a neural network to approximate the action-value function. ...
1
vote
1answer
53 views

Is the distribution of state-action pairs from sample based planning accurate for small experience sets?

From the David Silver's lecture 8: Integrating Learning and Planning - based on Sutton and Barto - he talks about using sample-based planning to use our model to take a sample of a state and then use ...
3
votes
1answer
34 views

Why is learning $s'$ from $s,a$ a kernel density estimation problem but learning $r$ from $s,a$ is just regression?

In David Silver's 8th lecture he talks about model learning and says that learning $r$ from $s,a$ is a regression problem whereas learning $s'$ from $s,a$ is a kernel density estimation. His ...
4
votes
2answers
333 views

Is there any good reference for double deep Q-learning?

I am new in reinforcement learning, but I already know deep Q-learning and Q-learning. Now, I want to learn about double deep Q-learning. Do you know any good references for double deep Q-learning? ...
1
vote
1answer
24 views

What are finite horizon look-ahead policies in reinforcement learning?

I was reading the paper How to Combine Tree-Search Methods in Reinforcement Learning published in AAAI Conference 2019. It starts with the sentence Finite-horizon lookahead policies are abundantly ...
2
votes
1answer
40 views

How should I decay $\epsilon$ in Q-learning?

How should I decay the $\epsilon$ in Q-learning? Currently, I am decaying epsilon as follows. I initialize $\epsilon$ to be 1, then, after every episode, I multiply it by some $C$ (let it be $0.999$)...
0
votes
0answers
69 views

Understanding the role of the target network in this DQN algorithm

I've found online this interesting algorithm: From what I understand reading this algorithm, I can't figure out why I should "perform the opposite action" and consequently storing that second ...
5
votes
1answer
125 views

Is this proof of $\epsilon$-greedy policy improvement correct?

The text book being referred to, in this question is "Reinforcement Learning: An introduction" by Richard Sutton and Andrew Barto (second edition, 2018). For your convenience, I have ...
1
vote
2answers
116 views

Is there any good source for when the pole actually starts all the way at the bottom, in the cartpole problem?

There are a lot of examples of balancing a pole (see image below) using reinforcement learning, but I find that almost all examples start close to the upright position. Is there any good source (or ...
0
votes
0answers
39 views

Using a model-based method to build an accurate day trading environment model

There are several different angles we can classify Reinforcement Learning methods from. We can distinguish three main aspects : Value-based and policy-based On-policy and off-policy Model-free and ...
0
votes
0answers
19 views

Can you find another reason for sample inefficiency of model-free on-policy Deep Reinforcement Learning?

The following mindmap gives an overview of multiple reasons for sample inefficiency. The list is definitely not complete. Can you see another reason not mentioned so far? Some related links: ...
1
vote
1answer
41 views

Why do we also need to normalize the action's values on continuous action spaces?

I was reading here tips & tricks for training in DRL and I noticed the following: always normalize your observation space when you can, i.e., when you know the boundaries normalize your ...
2
votes
0answers
38 views

Can weighted importance sampling be applied to off-policy evaluation for continuous state space MDPs?

Can weighted importance sampling (WIS) and importance sampling (IS) be applied to off-policy evaluation for continuous state spaces MDPs? Given that I have trajectories of $(s_t,a_t)$ pairs and the ...

1 2
3
4 5
25