Hottest Questions This Month

7

votes

3

answers

653

views

What are the purposes of autoencoders?

answered Mar 24 at 0:26 Josiah 141

8

votes

2

answers

82

views

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

reinforcement-learning q-learning deep-rl proofs function-approximation

answered Apr 6 at 9:16 Dennis Soemers 4,162

4

votes

1

answer

197

views

What loss function to use when labels are probabilities?

neural-networks machine-learning loss-functions probability-distribution

modified Apr 15 at 10:11 nbro 2,440

5

votes

2

answers

68

views

How do I improve accuracy and know when to stop training?

convolutional-neural-networks classification

modified Mar 26 at 7:59 mirror2image 313

3

votes

2

answers

126

views

Is there a concept of neural networks that edit neural networks?

neural-networks

answered Mar 28 at 23:51 John Doucette 3,465

4

votes

1

answer

87

views

What is a simplified way to explain why the AI researchers Bengio, Hinton, and Lecun, won the 2019 Turing Award?

neural-networks convolutional-neural-networks soft-question

modified Apr 3 at 18:15 Guillermo Mosse 167

8

votes

1

answer

109

views

What causes a model to require a low learning rate?

machine-learning hyper-parameters

modified Apr 2 at 3:19 Ben N♦ 2,020

4

votes

1

answer

59

views

Can this tic tac toe program be considered AI?

neural-networks game-ai

answered Apr 3 at 1:27 drerD 101

7

votes

1

answer

79

views

Are there reinforcement learning algorithms that scale to large problems?

reinforcement-learning

modified yesterday Philip Raeisghasem 1,125

4

votes

2

answers

96

views

Which loss function is the brain optimizing in order to learn advanced visual skills without expert/human supervision?

machine-learning computer-vision cognitive-science biology

modified Apr 11 at 21:03 Pablo Messina 47

5

votes

1

answer

95

views

When should we use algorithms like Adam as opposed to SGD?

machine-learning optimization

modified Mar 26 at 14:15 nbro 2,440

3

votes

3

answers

89

views

Can an artificial intelligence suffer?

philosophy emotional-intelligence artificial-life

modified Mar 31 at 12:43 nbro 2,440

4

votes

2

answers

86

views

Is decision tree learning a deterministic algorithm?

decision-tree random-variable

modified Apr 2 at 8:20 nbro 2,440

3

votes

2

answers

60

views

Is reinforcement learning using shallow neural networks still deep reinforcement learning?

deep-learning reinforcement-learning terminology deep-rl

modified Mar 31 at 19:30 Neil Slater 6,504

3

votes

1

answer

70

views

What is the difference between search and learning?

deep-learning philosophy search

modified Apr 14 at 22:58 John Doucette 3,465

1

vote

2

answers

67

views

Creative AI semester project (4 week time-frame)

ai-design game-ai ai-basics ai-community

modified Apr 4 at 17:50 DukeZhou♦ 4,107

4

votes

1

answer

59

views

Is it possible to make a 'forked path' neural network?

neural-networks machine-learning convolutional-neural-networks image-recognition

modified Apr 11 at 5:17 Sooryakiran Pallikulathil 76

0

votes

1

answer

75

views

Bachelor thesis in reinforcement learning

reinforcement-learning open-ai academia

modified Apr 13 at 20:45 John Doucette 3,465

5

votes

2

answers

93

views

Are on-line backpropagation iterations perpendicular to the constraint?

backpropagation math gradient-descent

modified Apr 7 at 20:52 IRFAN HAIDER 3

0

votes

1

answer

148

views

Can the inputs and outputs of a neural network be a neural network?

neural-networks

modified Mar 31 at 22:02 Neil Slater 6,504

3

votes

3

answers

88

views

Can AI research lead to new findings in general cognitive science?

philosophy cognitive-science

answered Apr 1 at 11:42 Douglas Daseeco 5,464

4

votes

1

answer

67

views

Unsupervised Learning for anomaly detection

neural-networks machine-learning deep-learning unsupervised-learning anomaly-detection

modified Mar 23 at 11:36 Pedro Henrique Monforte 191

5

votes

1

answer

52

views

Why didn't champion of the Go game manage to win the last game against AlphaGo, after winning the 4th one?

game-ai soft-question alphago monte-carlo

answered Mar 27 at 4:01 John Doucette 3,465

3

votes

2

answers

35

views

Shouldn't Gödel's Incompleteness Theorems disprove the Physical symbol system hypothesis?

philosophy symbolic-ai

modified Mar 29 at 14:47 John Doucette 3,465

6

votes

1

answer

44

views

Are there existing examples of using neural networks for static code analysis?

neural-networks machine-learning quality-control

answered Mar 25 at 15:19 John Doucette 3,465

4

votes

2

answers

44

views

What are the key differences between reinforcement learning & optimal control?

reinforcement-learning

answered Mar 22 at 16:25 Muhammad uxman 40

5

votes

1

answer

44

views

What can OpenCog do?

applications agi open-cog

answered Apr 6 at 23:41 hisairnessag3 897

4

votes

2

answers

56

views

Can DQN perform better than Double DQN?

reinforcement-learning q-learning keras dqn

answered Apr 9 at 7:35 mirror2image 313

3

votes

0

answers

115

views

How can I convert the problem formulation to multi-agent reinforcement learning?

reinforcement-learning models applications multi-agent-systems

modified Apr 8 at 14:01 i_th 116

2

votes

2

answers

65

views

Can Reinforcement Learning solve problems, where certain elements in the environement are randomly located?

deep-learning reinforcement-learning rewards

modified Mar 22 at 8:10 mirror2image 313

3

votes

2

answers

83

views

Do we also need to model a probability distribution for the decoder of a VAE?

computer-vision generative-model autoencoders latent-variable

modified Apr 2 at 9:10 DuttaA 2,505

1

vote

1

answer

49

views

How can the $\lambda$-return be defined recursively?

reinforcement-learning rl-an-introduction return eligibility-traces

modified Apr 14 at 7:44 Philip Raeisghasem 1,125

2

votes

2

answers

38

views

Is max pooling really bad?

neural-networks machine-learning deep-learning

answered Apr 5 at 5:55 hisairnessag3 897

0

votes

1

answer

76

views

Why isn't the reverse KL divergence commonly used in supervised learning?

machine-learning optimization loss-functions supervised-learning

modified Apr 8 at 8:19 user1587520 111

2

votes

1

answer

32

views

Can a neuron have both a bias and a threshold?

neural-networks neurons perceptron

modified Apr 6 at 19:55 Philip Raeisghasem 1,125

0

votes

1

answer

39

views

What does “Wide” vs. “Deep” mean in the context of Neural Networks?

neural-networks deep-learning

answered Mar 25 at 22:11 nbro 2,440

2

votes

2

answers

39

views

Do we have cross-language vector space for word embedding?

natural-language-processing word-embedding

answered yesterday HLeb 16

3

votes

2

answers

50

views

Can Q-learning be used in a POMDP?

reinforcement-learning q-learning pomdp mdp sarsa

modified Apr 3 at 11:43 nbro 2,440

1

vote

1

answer

38

views

What is the motivation behind using a deterministic policy?

reinforcement-learning deterministic-policy

modified Apr 4 at 19:32 Philip Raeisghasem 1,125

3

votes

1

answer

38

views

Why are not validation accuracy and loss as smooth as train accuracy and loss?

neural-networks convolutional-neural-networks training

modified Apr 10 at 17:57 DuttaA 2,505

1

vote

1

answer

48

views

Meaning of the statement $\forall x \exists y \forall z (z \neq y \iff f(x) \neq z)$

ai-basics logic

answered Apr 14 at 19:57 Philip Raeisghasem 1,125

2

votes

1

answer

47

views

Can I calculate the training performance of GPUs by comparing their specification?

training gpu

modified Apr 10 at 10:39 DuttaA 2,505

1

vote

2

answers

35

views

Can gamma be greater than 1 in a DQN?

q-learning dqn

answered Apr 8 at 6:39 i_th 116

2

votes

1

answer

23

views

Inform policy learning of environment constants

reinforcement-learning

modified Apr 5 at 15:03 Seanny123 111

0

votes

1

answer

48

views

Does backpropagation update weights one layer at a time?

neural-networks deep-learning

modified Apr 4 at 20:29 Philip Raeisghasem 1,125

2

votes

1

answer

48

views

How do I choose the number of neurons in the fully-connected layer before the softmax layer?

neural-networks deep-learning convolutional-neural-networks ai-design

modified Apr 16 at 9:33 nbro 2,440

3

votes

1

answer

37

views

How can alpha zero learn if the tree search stops and restarts before finishing a game?

deep-learning reinforcement-learning monte-carlo-tree-search alphazero

modified Apr 12 at 19:07 Philip Raeisghasem 1,125

0

votes

2

answers

37

views

How can the A* algorithm be optimized?

algorithm optimization search a-star

modified Apr 16 at 15:33 nbro 2,440

Top Questions

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