The Stack Overflow podcast is back! Listen to an interview with our new CEO.

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
1
vote
1answer
19 views

Confusion about temporal difference learning

I have a couple of small questions about the David Silver lecture about reinforcement learning, lecture slides (slides 23, 24). More specifically it is about the temporal difference algorithm: $$V(...
1
vote
0answers
10 views

Is there any open source implementation of the SBEED learning algorithm?

Are there are any openly available implementations of the SBEED: Convergent Reinforcement Learning with Nonlinear Function Approximation learning?
1
vote
0answers
25 views

Reinforcement learning number of episodes per epoch not matching with paper

I am trying to reproduce results presented in this paper. On page 4, the authors state: ... we train for 50 epochs (one epoch consists of 19*2*50 = 1900 full episodes), which amounts to a total ...
1
vote
1answer
71 views

An “elevator pitch” breakdown of areas of applications for Reinforcement Learning & Neural Networks vs. Genetic Algorithms

I'm looking for an "elevator pitch" breakdown of areas of applications for Reinforcement Learning & Neural Networks vs. Genetic Algorithms, both actual and theoretical. Links are welcome, but ...
2
votes
1answer
25 views

Doubt in Deep-Q learning with sparse rewards

I am working on a deep reinforcement learning problem, when I got stuck at the following questions. They are rather general and not specific to my specific problem. The solution uses a sparse reward ...
2
votes
1answer
26 views

Q Learning for FrozenLake environment not converging to V* values from Value Iteration

I am trying to learn tabular Q learning, value iteration using the classical algorithms (no neural networks) by using a table of states and actions. I was trying it out on FrozenLake environment in ...
2
votes
1answer
34 views

What are sim2sim, sim2real and real2real?

Recently, I always hear about the terms sim2sim, sim2real and real2real. Will anyone explain the meaning/motivation of these terms (in DL/RL research community)? What are the challenges in this ...
0
votes
0answers
18 views

Ideas on how to train an AI to play Mario Kart with the DeSmuME Emulator

Gday guys, i have this idea in my mind for quite a while. I want to teach an AI to play Mario Kart on the NDS, which can hopefully beat me and my friends one day. Iam familiar with the theoretical ...
2
votes
0answers
28 views

Best approach for online Machine Translation with few hundred of samples?

I want to implement a model that improves itself with the passage of time. My main task is to build a machine translator (from English to Urdu).. The problem I am facing is that I have very little ...
0
votes
1answer
55 views

Model-based Reinforcement Learning algorithm for real-time robotics task

I'm quite a newbie when it comes to practically working with Deep Learning techniques, although I studied them quite a lot theoretically in the last months. However, now I'm facing my first practical ...
2
votes
1answer
43 views

How can I teach a computer to play N64 games using Neural Nets?

I would like to work on a project where I teach an NN to play N64 games. To my current understanding, I would need an emulator? I can do the Machine Learning side of it, im just unsure how I can ...
3
votes
0answers
46 views

Sampling in TRPO or PPO

In the TRPO paper, the objective to maximize is (equation 14) $$ \mathbb{E}_{s\sim\rho_{\theta_\text{old}},a\sim q}\left[\frac{\pi_\theta(a|s)}{q(a|s)} Q_{\theta_\text{old}}(s,a) \right] $$ which ...
2
votes
0answers
23 views

Is there a way to do reinforcement learning in POMDP?

Are there any algorithms to use reinforcement learning to learn optimal policies in partially observable Markov decision process (POMDP) i.e. when the state is not perfectly observed. More ...
0
votes
0answers
12 views

Choosing best combinations from all possible combination expressions based few variables, unary operators, binary operators

I have a few financial variables of a stock universe like OHLC prices, volume, and other fundamentals with varying time-frequency. Using this set I'm creating an expression that gives the weights to ...
1
vote
1answer
29 views

Can reinforcement learning be utilized for creating a simulation?

According to the definition, the AI agent has to play a game by it's own. A typical domain is the blocksworld problem. The AI determines which action the robot in a game should execute and a possible ...
1
vote
1answer
41 views

What is the difference between the definition of a stationary policy in reinforcement learning and contextual bandit?

A stationary policy is a function that maps a state to a probability distribution of actions. In a contextual bandit problem, a state itself does not include the history. But in a reinforcement ...
2
votes
1answer
52 views

Markov property in maze solving problem in reinforcement learning

By definition, every state in RL has Markov property, which means that the future state depends only on the current state, not the past states. However, I saw that in some case we can define a state ...
4
votes
1answer
42 views

How would one implement a multi-agent environment with asynchronous action and rewards per agent?

In a single agent environment, the agent takes an action, then observes the next state and reward: ...
1
vote
1answer
33 views

Unable to replicate Figure 2.1 from “Reinforcement Learning: An Introduction”

The author explains in 2.2 Action-Value Methods: To roughly assess the relative effectiveness of the greedy and $\varepsilon $-greedy methods, we compared them numerically on a suite of test ...
3
votes
1answer
31 views

Why does having a fixed policy change a Markov Decision Process (MDP) to a Markov Reward Process (MRP)?

If a policy is fixed, it is said that an MDP becomes an MRP. Why is this so? Aren't the transitions and rewards still parameterized by the action and current state? In other words, aren't the ...
4
votes
2answers
55 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 ...
2
votes
0answers
25 views

how to use Softmax action selection algorithm in atari-like game

I'm currently writing a program using keras (python 3) to play a game similar to Atari games, only in this one there are objects moving in the screen in different angles and directions (in most of ...
1
vote
0answers
23 views

RF: How to deal with environments changing state due to external factors

I have a use case where the state of the environment could change due to random events in between time steps that the agent takes actions. For example, at t1, the agent takes action a1 and is given ...
1
vote
0answers
12 views

TD losses are descreasing, but also rewards are decreasing, increasing sigma

I'm using Q-learning with some extensions such as noisy linear layers, n-steps and double DQN. The training, however, isn't that successful, my rewards are descreasing over time after a steep ...
5
votes
1answer
34 views

Is the agent aware of a possible different set of actions for each state?

I have a use case where the set of actions is different for different states. Is the agent aware of what actions are valid for which states or is the agent only aware of the entire action space (in ...
0
votes
0answers
18 views

DQN unlearns certain OpenAI-Gym environments

I solved the OpenAI-Gym MountainCar-v0 environment using dqn(using low-state-dimensional input). When I used the same code for solving CartPole-v0 environment, the network got trained in the reverse ...
3
votes
0answers
34 views

What is the difference between random and sequential sampling from the reply memory?

I was working on an RL problem and I am confused at one specific point. We use replay memory so that the network learns about previous actions and how these actions lead to a success or a failure. ...
0
votes
1answer
33 views

Why doesn't stability in prediction imply stability in control in off-policy reinforcement learning?

Prediction's goal is to get an estimate of a performance of a policy given a specific state. Control's goal is to improve the policy wrt. the prediction. The alternation between the two is the ...
3
votes
1answer
38 views

What are the differences between stability and convergence in reinforcement learning?

The terms are mentioned in the paper: “An Emphatic Approach to the Problem of off-Policy Temporal-Difference Learning.” (Sutton, Mahmood, White; 2016) and more, of course. In which paper, they ...
-1
votes
1answer
25 views

Does higher Accuracy in Reinforcement Learning indicate better model performance?

If a reinforcement learning algorithm uses a Deep Neural Network to predict the action given a state (a NN for a policy function), an Monte Carlo Tree Search in a model-based learning setup, then ...
0
votes
0answers
18 views

A2C for the game of Hanabi underfits

I am trying to solve the game of Hanabi (paper describing game) with actor-critic algorithm. I took code for the environment from the Deepmind's repository and implemented a2c algorithm myself. From ...
2
votes
1answer
38 views

How do policy gradients compute an infinite probability distribution from a neural network

Do neural networks compute the probability distribution for policy gradient methods. If so, how do they compute an infinite probability distribution? How do you represent a continuous action policy ...
2
votes
1answer
46 views

What happens to the optimal value function if the reward is multiplied by a constant?

What happens to the optimal action-value function, $q_*$ if the reward is multiplied by a constant $c$? Is the optimal action-value function also multiplied by such a constant?
2
votes
1answer
62 views

Monte Carlo learning for Reinforcement learning

When you train a model using Monte Carlo-based learning the state and action taken at each step is recorded, and then at some point an end state is reached and the agent receives some reward - what do ...
2
votes
1answer
47 views

Can an RL algorithm trained in one environment be successful in a different one?

Can an RL algorithm trained in one environment be successful in a different one? For example, if I train a model to go through one labyrinth, could this model also go through a different but similar ...
0
votes
1answer
64 views

How can the derivative of a neural network be calculated, given no mathematical expression?

Neural networks (NNs) are used as approximators in reinforcement learning (RL). To update the policy in RL, the actor network's gradients w.r.t its weights are needed. Since NN doesn't have a ...
5
votes
1answer
35 views

Reinforcement learning with incremental action space

Let's say we have a problem which can be solved by some RL algorithms (DQN, for example, because we have discrete action space). At first, the action space is fixed (the number of actions is $n_1$), ...
1
vote
1answer
33 views

How is the actor-critic algorithm guaranteed to converge?

From my understanding, the critic evaluates the policy (actor) following dynamic programming (DP) or approximate dynamic programming (ADP) scheme, which should converge to the optimal value function ...
1
vote
0answers
50 views

Is Value Iteration better than Policy Iteration for first few iterations?

In Policy Iteration (PI), the action generated by the policy, whether it's optimal or not w.r.t the current value function $v(s)$. Whereas, in Value Iteration, the action is greedily generated w.r.t ...
1
vote
1answer
29 views

Can policy iteration use only the immediate reward for updates?

Is it still a policy iteration algorithm if the policy is updated optimizing a function of the immediate reward instead of the value function?
3
votes
1answer
32 views

How important is the choice of the initial state?

Is it crucial to always have the same initial (starting) state for Reinforcement Learning, for example, for Q-learning or DQN? Or it can vary?
3
votes
1answer
54 views

What is a non-starving policy in reinforcement learning?

In the paper, Eligibility Traces for off-Policy Policy Evaluation (2010), by Doina Precup et al., mentioned the term "non-starving" many times. The specific use of the term was like "non-starving ...
3
votes
1answer
31 views

Reinforcement learning with hints or reference model

In Reinforcement Learning, when I train a model, it comes up with its own set of solutions. For example, if I am training a robot to walk, it will come up with its own walking gait, such as this Deep ...
0
votes
0answers
14 views

Invalid moves in Deep Reinforcement Learning for games [duplicate]

I've been working on a bot for a game involving dice throws and chance. The architecture involved is similar to AlphaZero in the that it has Convolutions and MCTS. According to the current state ...
1
vote
0answers
22 views

PPO: action std or entropy for exploration?

When trying to implement my own PPO (Proximal Policy Optimizer), I came accross two different implementations : Exploration with action std : Collect trajectories on ...
3
votes
1answer
72 views

How is the gradient of the loss function in DQN derived?

In the original DQN paper, page 1, the loss function of the DQN is $$ L_{i}(\theta_{i}) = \mathbb{E}_{(s,a,r,s') \sim U(D)} [(r+\gamma \max_{a'} Q(s',a',\theta_{i}^{-}) - Q(s,a;\theta_{i}))^2] $$ ...
2
votes
1answer
51 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 ...
0
votes
0answers
30 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. ...
4
votes
1answer
67 views

Benchmarks for reinforcement learning in discrete MDPs

To compare the performance of various algorithms for perfect information games, reasonable benchmarks include reversi and m,n,k-games (generalized tic-tac-toe). For imperfect information games, ...
2
votes
1answer
42 views

Can I have different rewards for a single action based on which state it transitions to?

I am working on an MDP where there are four states and ten actions. I am supposed to derive the optimal policy to reach the desired state. At any state, a particular action can take you to any of the ...