Questions tagged [reinforcement-learning]

For questions related to reinforcement learning, i.e. a machine learning technique where we imagine an agent that interacts with an environment (composed of states) in time steps by taking actions and receiving rewards (or reinforcements), then, based on these interactions, the agent tries to find a policy (i.e. a behavioural strategy) that maximizes the cumulative reward (in the long run), so the goal of the agent is to maximize the reward.

Filter by
Sorted by
Tagged with
4
votes
1answer
67 views

REINFORCE algorithm for portfolio optimization - problem while training

I'm trying to implement the Reinforce algorithm (Monte Carlo policy gradient) in order to optimize a portfolio of 94 stocks on a daily basis (I have suitable historical data to achieve this). The idea ...
0
votes
1answer
64 views

How can I model this problem of delivering assets by choosing a route with reinforcement learning?

I would like to build a model based on reinforcement learning (RL) for the following scenario Recommend the best route (of cities listed for a given country) that satisfies the required criteria (...
2
votes
1answer
93 views

Why is the expected return in Reinforcement Learning (RL) computed as a sum of cumulative rewards?

Why is the expected return in Reinforcement Learning (RL) computed as a sum of cumulative rewards? Would it not make more sense to compute $\mathbb{E}(R \mid s, a)$ (the expected return for taking ...
2
votes
0answers
45 views

What kind of policy evaluation and policy improvement AlphaGo, AlphaGo Zero and AlphaZero are using

I'm trying to find out what kind of policy improvement and policy evaluation AlphaGo, AlphaGo Zero, and AlphaZero are using. By looking into their respective paper and SI, I can conclude that it is a ...
2
votes
0answers
49 views

Prioritised Remembering in Experience Replay (Q-Learning)

I'm using Experience Replay based on the original Prioritized Experience Replay (PER) paper. In the paper authors show ~ an order of magnitude increase in data efficiency from prioritized sampling. ...
2
votes
1answer
223 views

Deep Reinforcement Learning: Rewards suddenly dip down

I am working on a deep reinforcement learning problem. The policy network has the same architecture as the one Deepmind published in 'Playing Atari with Deep Reinforcement Learning'. I am also using ...
1
vote
1answer
178 views

Where are reinforcement algorithms used in financial services?

One of the most common misconceptions about reinforcement learning (RL) applications is that, once you deploy them, they continue to learn. And, usually, I'm left having to explain this. As part of my ...
0
votes
1answer
67 views

What is the purpose of a Neural Network in Reinforcement Learning when we have a Q-learning update rule?

I'm confused as to the purpose of training a neural network (NN) for reinforcement learning (RL) tasks such as Gridworld. In RL tasks, namely q-learning, we have a q-learning update rule, which is ...
1
vote
0answers
44 views

Correct dimensionality of parameter vector for solving an MRP with linear function approximation?

I'm in the process of trying to learn more about RL by shadowing a course offered collaboratively by UCL and DeepMind that has been made available to the public. I'm most of the way through the course,...
0
votes
1answer
66 views

Soft Actor Critic - Losses are not converging

I'm trying to implement soft actor-critic algorithm for financial data (stock prices) and I have trouble with losses, no matter what combination of HyperParameters I enter, they are not converging, ...
2
votes
1answer
40 views

What is convergence analysis, and why is it needed in reinforcement learning?

While reading a paper about Q-learning in network energy consumption, I came across the section on convergence analysis. Does anyone know what convergence analysis is, and why is convergence analysis ...
1
vote
0answers
35 views

How can I convert a simple CLI RPG to a compatible environment for training an RL agent via stable-baselines?

What would be the good choice of algorithm to use for character action selection in an RPG, implemented in Python? I had previously asked this question in the hopes of getting headway on the AI ...
4
votes
1answer
72 views

What does the term $|\mathcal{A}(s)|$ mean in the $\epsilon$-greedy policy?

I've been looking online for a while for a source that explains these computations but I can't find anywhere what does the $|A(s)|$ mean. I guess $A$ is the action set but I'm not sure about that ...
0
votes
0answers
29 views

Training a neural network for games with dynamic starting locations

I've been training a snake for the snake game with Dueling Double DQN as well as the Double DQN. I've noticed that I can see progress being made when the snake starts off from the same location every ...
1
vote
1answer
37 views

How to optimize neural network parameters with REINFORCE

I've seen a few mentions in papers that neural network parameters can be found using REINFORCE algorithm. It was mentioned in the context of nondifferentiable operations involving e.g. step function ...
3
votes
1answer
108 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: "...
2
votes
0answers
43 views

How can I formulate a prediction problem (given labeled data) as an RL problem and solve it with Q-learning?

One of my friends sent me a problem he was working on lately, and I couldn't help but I wonder how could it be solved using Q-learning. The statement is as follows: Given the following datasets, the ...
1
vote
2answers
486 views

Is Monte Carlo Tree Search appropriate for problems with large state and action spaces?

I'm doing a research on a finite-horizon Markov decision process with $t=1, \dots, 40$ periods. In every time step $t$, the (only) agent has to chose an action $a(t) \in A(t)$, while the agent is in ...
2
votes
1answer
51 views

How can I classify policy gradient methods in RL?

In the book of Barto and Sutton, there are 3 methods presented that solve an RL problem: DP, Monte Carlo, and TD. But which category does policy gradient methods (or actor-only methods) classify in? ...
1
vote
1answer
51 views

Why does adding another network help in double DQN?

What is the idea behind double DQN? The Bellman equation used to calculate the Q values to update the online network follows the equation: ...
1
vote
1answer
57 views

Do smaller loss values during DQN training produce better policies?

During the training of DQN, I noticed that the model with prioritized experience replay (PER) had a smaller loss in general compared to a DQN without PER. The mean squared loss was an order of ...
0
votes
1answer
59 views

How to “forward” updated NN model to a transferred model?

I've trained a robot to walk in a straight line for as long as it can (using TD3), and now I'm using that pre-trained model for two new models with separate purposes: 1. Walk to a specific point and ...
1
vote
1answer
54 views

Can rewards be decomposed into components?

I'm training a robot to walk to a specific $(x, y)$ point using TD3, and, for simplicity, I have something like ...
1
vote
1answer
34 views

If the transition model is available, why would we use sample-based algorithms?

Sample-based algorithms, like Monte Carlo Algorithms and TD-Learning, are often presented as useful since they do not require a transition model. Assuming I do have access to a transition model, are ...
1
vote
2answers
86 views

Running 2 NEAT nets on the same observations

So i have been playing around with neat-python. I made a program, applying neat, to play pinball on the Atari 2600. The code for that can be found in the file ...
1
vote
0answers
29 views

Solving multi-armed bandit problems with continuous action space

My problem has a single state and an infinite amount of actions on a certain interval (0,1). After quite some time of googling I found a few paper about an algorithm called zooming algorithm which can ...
15
votes
3answers
993 views

Why doesn't Q-learning converge when using function approximation?

The tabular Q-learning algorithm is guaranteed to find the optimal $Q$ function, $Q^*$, provided the following conditions (the Robbins-Monro conditions) regarding the learning rate are satisfied $\...
0
votes
1answer
530 views

Reinforcement learning with PPO: rewards decreasing

I'm trying to train a PPO agent and my average rewards graph looks like this. Could this indicate that it's stuck at a local maximum? Do I need to promote exploring by increasing the entropy or does ...
1
vote
1answer
91 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 ...
3
votes
2answers
114 views

What is the difference between return and expected return?

At a time step $t$, for a state $S_{t}$, the return is defined as the discounted cumulative reward from that time step $t$. If an agent is following a policy (which in itself is a probability ...
2
votes
1answer
81 views

How is the reward in reinforcement learning different from the label in supervised learning problems?

How is the notion of immediate reward used in the reinforcement learning different from the notion of a label we find in the supervised learning problems?
4
votes
2answers
164 views

Why AlphaGo didn't use Deep Q-Learning?

In the previous research, in 2015, Deep Q-Learning shows its great performance on single player Atari Games. But why do AlphaGo's researchers use CNN + MCTS instead of Deep Q-Learning? is that because ...
5
votes
1answer
118 views

Can supervised learning be recast as reinforcement learning problem?

Let's assume that there is a sequence of pairs $(x_i, y_i), (x_{i+1}, y_{i+1}), \dots$ of observations and corresponding labels. Let's also assume that the $x$ is considered as independent variable ...
1
vote
0answers
23 views

Should the importance sampling ratio be updated at the end of the for loop in the off-policy Monte Carlo control algorithm?

I'm studying RL with Sutton and Barto's book. I'd like to ask about the order of execution of a statement in the algorithm below. Here, $W$ (importance sampling ratio) is updated at the end of the <...
0
votes
1answer
46 views

How does training for DQN work if messing up in the environment in costly?

Suppose that we want to train a car to drive in the real world and decide to use Reinforcement Learning (specifically, DQN) for that. I am a bit confused about how training generally works. Is it that ...
6
votes
1answer
2k views

Suitable reward function for trading buy and sell orders

I am working to build an deep reinforcement learning agent which can place orders (i.e. limit buy and limit sell orders). The actions are ...
4
votes
1answer
270 views

Why does the DQN not converge when the start or goal states can change dynamically?

I'm trying to apply a DQN to a stochastic environment, but I'm having trouble getting it to converge. I found some similar questions asked here, but no solutions yet. I can easily get the DQN to ...
2
votes
1answer
52 views

Why do some DQN implementations not require random exploration but instead emulate all actions?

I've found online some DQN algorithms that (in a problem with a continuous state space and few actions, let's say 2 or 3), at each time step, compute and store (in the memory used for updating) all ...
1
vote
1answer
43 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 ...
1
vote
0answers
36 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 ...
2
votes
1answer
79 views

Into which subcategories can reinforcement learning be divided?

In the course of a scientific work, I will discuss the different types of reinforcement learning. However, I have difficulties to find these different types. So, into which subcategories can ...
0
votes
0answers
41 views

What does the notation “for t=T to 1,−1 do” in terms of time steps, in deep recurrent q network?

In looking at an algorithm in the paper Learning to Communicate with Deep Multi-Agent Reinforcement Learning. Here is the full algorithm: What does the notation ...
1
vote
0answers
31 views

What is the role of embeddings in a deep recurrent Q network?

When describing the model architecture for a deep recurrent q network, the authors of the paper Learning to Communicate with Deep Multi-Agent Reinforcement Learning each agent consists of a recurrent ...
2
votes
1answer
85 views

Reinforcement Learning Continuous Control (DDPG): How to avoid thrashing of issued actions? How to reward smooth output over flittering?

Currently I'm working on a continuous state / continuous action controller. It shall control a certain roll angle of an aircraft by issuing the correct aileron commands (between -1...1 continuous). ...
3
votes
2answers
1k views

What should I do when the potential value of a state is too high?

I'm working on a Reinforcement Learning task where I use reward shaping as proposed in the paper Policy invariance under reward transformations: Theory and application to reward shaping (1999) by ...
0
votes
1answer
45 views

What reinforcement learning algorithm should I use in continuous states?

I want to use reinforcement learning in an environment I made. The exact environment doesn't really matter, but it comes down to this: The amount of different states in the environment is infinite e.g....
5
votes
1answer
102 views

How do I convert an MDP with the reward function in the form $R(s,a,s')$ to and an MDP with a reward function in the form $R(s,a)$?

The AIMA book has an exercise about showing that an MDP with rewards of the form $r(s, a, s')$ can be converted to an MDP with rewards $r(s, a)$, and to an MDP with rewards $r(s)$ with equivalent ...
2
votes
1answer
81 views

How can I perform lane detection with reinforcement learning?

I'm quite new to reinforcement learning and my project will consist of detecting lanes with RL. I'm using q-learning and I'm having a hard time thinking how my q table should look like, I mean - ...
0
votes
0answers
29 views

DQN Tic-Tac-Toe does not quite become optimal

I am trying to implement a DQN agent for playing standard 3x3 Tic-Tac-Toe (it is a double DQN with experience replay, and using a target network). I got the hyperparameters to the point where the ...
1
vote
0answers
29 views

How are the coefficients of the Region of Interest being selected?

I was reading the following paper: Rl-Ncs: Reinforcement Learning Based Data-Driven Approach For Nonuniform Compressed Sensing, and my question is: how do they decide whether a signal is characterized ...

1
3 4
5
6 7
28