Hottest Questions This Month

5

votes

4

answers

542

views

Is the optimal policy always stochastic if the environment is also stochastic?

answered Feb 16 at 2:37 ron dembo 1

6

votes

3

answers

131

views

How can a neural network learn when the derivative of the activation function is 0?

neural-networks deep-learning gradient-descent

modified Jan 28 at 17:13 Douglas Daseeco 1

3

votes

2

answers

76

views

Is it possible to use AI to denoise noised or 'dirty' documents?

machine-learning ai-community ocr

answered Jan 24 at 18:04 han_nah_han_ 255

2

votes

3

answers

112

views

Are there profitable hedge funds using AI?

machine-learning research algorithmic-trading

answered Feb 15 at 14:18 KitsuneMakai 21

3

votes

2

answers

87

views

Rollout algorithm like Monte Carlo search suggest model based reinforcement learning?

reinforcement-learning models monte-carlo-tree-search

modified Feb 8 at 21:37 Manuel Rodriguez 1,396

3

votes

4

answers

85

views

Choose model with the smallest loss or highest accuracy?

neural-networks machine-learning academia

modified yesterday malioboro 458

1

vote

2

answers

102

views

Is the expert system still in use today?

ai-basics philosophy expert-system

modified Feb 5 at 23:22 malioboro 458

-3

votes

1

answer

208

views

The challenges of exports,imports,AI related [closed]

philosophy ethics autonomous-weapons cognition

modified Feb 15 at 13:29 ABHIJIT BAGCHI 1

1

vote

2

answers

93

views

Why does is make sense to normalize rewards per episode in reinforcement learning?

machine-learning reinforcement-learning

modified Jan 26 at 12:55 Dennis Soemers 3,797

2

votes

2

answers

70

views

Each training run for DDQN agent takes 2 days, and still ends up with -13 avg score, but OpenAi baseline DQN needs only an hour to converge to +18?

deep-learning reinforcement-learning dqn open-ai deepmind

answered Feb 10 at 3:56 hridayns 131

1

vote

2

answers

81

views

Is there more than one Q-matrix update formula?

reinforcement-learning python q-learning

modified Feb 18 at 20:46 Dennis Soemers 3,797

2

votes

2

answers

77

views

When is knowledge base consistent?

logic knowledge-representation ai-a-modern-approach

modified Feb 11 at 21:05 nbro 1,292

3

votes

3

answers

73

views

Status of discussion forum discuss.openai.com in 2019

ai-community

modified Jan 30 at 17:43 Neil Slater 5,468

3

votes

4

answers

112

views

Could artificial general intelligence harm humanity?

control-problem superintelligence mythology-of-ai neo-luddism value-alignment

answered Feb 11 at 10:08 Manuel Rodriguez 1,396

2

votes

2

answers

78

views

How to use CNN for making predictions on non-image data?

convolutional-neural-networks tensorflow

modified Feb 8 at 22:08 malioboro 458

1

vote

2

answers

41

views

How are the reward functions $R(s)$, $R(s, a)$ and $R(s, a, s')$ equivalent?

reinforcement-learning definitions markov-decision-process proofs

modified Feb 7 at 23:01 nbro 1,292

3

votes

3

answers

142

views

Can the rewards be stochastic when the transition model is deterministic?

reinforcement-learning markov-decision-process

modified Feb 13 at 7:43 Neil Slater 5,468

2

votes

1

answer

51

views

What information should be cached in experience replay for actor-critic?

machine-learning reinforcement-learning policy-gradients experience-replay

modified Feb 6 at 14:41 Gulzar 198

1

vote

2

answers

84

views

How is it that AI can become biased, and what are the proposals to mitigate this?

modified Feb 17 at 21:50 Nyos 31

1

vote

2

answers

62

views

Choosing the right neural network settings

neural-networks machine-learning convolutional-neural-networks

answered Jan 31 at 15:23 Edward Dixon 309

0

votes

1

answer

62

views

Difference between Machine learning and Deep learning? [closed]

machine-learning deep-learning

answered Feb 4 at 14:34 Mahmoud 49

1

vote

2

answers

61

views

RL vs Supervised Learning vs Planning

machine-learning reinforcement-learning planning

modified yesterday SeeDerekEngineer 118

3

votes

2

answers

70

views

Are neural networks statistical models?

neural-networks machine-learning deep-learning statistical-ai

modified Jan 29 at 17:02 Douglas Daseeco 1

3

votes

2

answers

69

views

What is the difference between search and planning?

terminology search definitions planning

modified Jan 28 at 12:56 nbro 1,292

0

votes

1

answer

87

views

What is the most biologically plausible representation for the actor and critic?

neural-networks reinforcement-learning actor-critic

answered Feb 15 at 8:32 yewang 36

1

vote

1

answer

52

views

How do updates in SARSA and Q-learning differ in code?

reinforcement-learning q-learning

modified Feb 9 at 18:55 Brale_ 311

1

vote

2

answers

59

views

How do we create a good agent that does not outperform humans?

reinforcement-learning philosophy intelligent-agent

modified Feb 15 at 17:37 nbro 1,292

2

votes

1

answer

59

views

What is the next state for a two-player board game?

reinforcement-learning q-learning

modified Feb 10 at 21:18 nbro 1,292

2

votes

2

answers

75

views

What is self-supervision in machine learning?

machine-learning supervised-learning self-supervised-learning

modified 9 hours ago nbro 1,292

1

vote

2

answers

57

views

How do self-driving cars construct paths?

self-driving path-planning pathfinding

modified Feb 13 at 15:04 Oliver Mason 1,068

2

votes

1

answer

56

views

Will Q-learning converge to the optimal state-action function when the reward periodically changes?

reinforcement-learning q-learning

modified Feb 15 at 17:40 nbro 1,292

2

votes

1

answer

56

views

Calculating gradient for log policy when variance is not constant

reinforcement-learning policy-gradients

answered Feb 15 at 9:47 yewang 36

-2

votes

0

answers

92

views

Should Facebook's Safety Check A.I. work if an account is stuck at a Name Change Checkpoint page? [closed]

modified Feb 15 at 6:58 prosody-Gab Vereable Context 97

2

votes

2

answers

55

views

Using GAN's to generate dataset for CNN training

convolutional-neural-networks datasets generative-adversarial-networks

answered Jan 31 at 13:03 Edward Dixon 309

1

vote

1

answer

70

views

How do I convert table-based to neural network-based Q-learning?

neural-networks reinforcement-learning q-learning

modified Feb 10 at 20:27 nbro 1,292

1

vote

1

answer

54

views

What is the relation between online learning and on-policy algorithms?

reinforcement-learning terminology definitions on-policy off-policy

modified Feb 10 at 17:27 Neil Slater 5,468

2

votes

3

answers

67

views

Does leaky relu help learning if the final output needs negative values?

neural-networks

modified Feb 4 at 8:47 DuttaA 2,141

2

votes

2

answers

66

views

Is Windows a bad choice for DRL?

reinforcement-learning software-evaluation

modified 2 days ago DukeZhou♦ 3,777

1

vote

1

answer

51

views

What are temporal-difference and Monte Carlo methods intuitively?

reinforcement-learning monte-carlo temporal-difference

modified Feb 15 at 14:34 nbro 1,292

2

votes

1

answer

63

views

Can TD-Lambda be used with Deep Reinforcement Learning?

neural-networks machine-learning reinforcement-learning

modified Feb 3 at 23:23 nbro 1,292

-2

votes

1

answer

101

views

What is the difference between Deep learning and Machine learning? [closed]

machine-learning deep-learning

modified Jan 28 at 16:59 Douglas Daseeco 1

3

votes

3

answers

46

views

What does learning mean?

machine-learning philosophy theory

answered Jan 30 at 15:01 Pravin RGMishra 1

2

votes

2

answers

46

views

Does fp32 & fp64 performance of GPU affect deep learning model training?

deep-learning hardware hardware-evaluation gpu

modified Feb 14 at 9:03 ssegvic 259

1

vote

1

answer

44

views

Meaning of Actor Output in Actor Critic Reinforcement Learning

machine-learning reinforcement-learning q-learning policy-gradients actor-critic

answered Feb 6 at 18:07 Dennis Soemers 3,797

3

votes

1

answer

44

views

Can Q-learning be used to derive a stochastic policy?

reinforcement-learning q-learning stochastic-policy deterministic-policy

modified Feb 8 at 14:45 Hammer. Wang 18

1

vote

1

answer

43

views

Is my Neural Network program fully connected?

neural-networks machine-learning ai-design lstm

modified Jan 28 at 18:23 dexteritas 138

1

vote

1

answer

54

views

A neural network for digits recognition doesn't work (MNIST, Numpy) [closed]

neural-networks machine-learning backpropagation object-recognition perceptron

answered Feb 10 at 22:31 Андрій Немченко 16

1

vote

2

answers

42

views

When are Q values calculated in experince replay?

machine-learning reinforcement-learning q-learning

modified Feb 2 at 9:45 Neil Slater 5,468

Top Questions

Looking for more? Browse the complete list of questions, or popular tags. Help us answer unanswered questions.