User OmG

5 votes

Accepted

What does "semantic gap" mean?

answered Apr 18, 2021 at 4:52

4 votes

Accepted

What is numerical stability?

answered Aug 3, 2021 at 14:11

4 votes

Accepted

Why are Siamese Neural Networks used instead of a single neural network?

answered Apr 12, 2022 at 20:11

4 votes

Can prior knowledge be encoded in deep neural networks?

answered Feb 15, 2018 at 17:21

3 votes

Accepted

How can active learning be used in the case of complex models that require a lot of data?

answered Dec 6, 2018 at 15:09

3 votes

What's going on in the equation of the variational lower bound?

answered Oct 21, 2019 at 11:49

3 votes

Accepted

How does the Kullback-Leibler divergence give "knowledge gained"?

answered Oct 22, 2019 at 8:10

3 votes

Accepted

How does size of the dataset depend on VC dimension of the hypothesis class?

answered Apr 17, 2020 at 0:26

3 votes

Accepted

What is the difference between hill-climbing and greedy best-first search algorithms?

answered Apr 24, 2020 at 23:11

3 votes

Accepted

What are the typical sizes of practical/commercial artificial neural networks?

answered May 16, 2021 at 16:18

3 votes

Can Reinforcement Learning be used to generate sequences?

answered May 26, 2021 at 18:48

3 votes

Does elitism cause premature convergence in genetic algorithms?

answered May 28, 2021 at 0:37

3 votes

What are the main ideas behind NER?

answered Jun 15, 2020 at 10:11

3 votes

Accepted

Any comparison between transformer and RNN+Attention on the same dataset?

answered Oct 5, 2020 at 9:19

2 votes

Accepted

Is the VC Dimension meaningful in the context of Reinforcement Learning?

answered Jun 19, 2020 at 12:00

2 votes

Which metric should I use to assess the quality of the clusters?

answered Feb 9, 2021 at 21:28

2 votes

Is knowing underlying probability distribution mandatory for deciding iid property of random variables?

answered Sep 13, 2021 at 17:52

2 votes

What's the threshold to call something 'machine learning'?

answered Sep 24, 2020 at 13:34

2 votes

Accepted

What does "Gau" in GauGAN stand for?

answered Dec 12, 2021 at 11:18

2 votes

Are there deep networks that can differentiate object class from individual object?

answered Oct 15, 2019 at 9:27

2 votes

Accepted

Understanding relation between VC Symmetrization Lemma and Generalization Bounds

answered Apr 14, 2020 at 18:59

2 votes

What kind of problems cannot be solved using machine learning techniques?

answered Jun 5, 2020 at 16:33

2 votes

Why does estimation error increase with $|H|$ and decrease with $m$ in PAC learning?

answered Sep 16, 2019 at 11:25

2 votes

Accepted

How do I prove that $\mathcal{H}$, with $\mathcal{VC}$ dimension $d$, shatters all subsets with size less than $d-1$?

answered Mar 29, 2020 at 0:29

2 votes

Accepted

What does the notation ${s'\sim T(s,a,\cdot)}$ mean?

answered Mar 29, 2020 at 16:06

2 votes

What is the name of an AI system that learns by trial and error?

answered Oct 3, 2018 at 16:57

2 votes

Why does the hill climbing algorithm only produce a local maximum?

answered Nov 27, 2018 at 9:29

2 votes

Why does Google Translate produce two different translations for the same Chinese text?

answered Jan 13, 2018 at 21:27

2 votes

Accepted

What is the difference between meta-learning and zero-shot learning?

answered Jan 23, 2019 at 16:58

2 votes

Accepted

Do I need an encoder-decoder architecture to predict the next item of a sequence?

answered Feb 16, 2019 at 16:51

Stack Exchange Network

75 Answers

What does "semantic gap" mean?

What is numerical stability?

Why are Siamese Neural Networks used instead of a single neural network?

Can prior knowledge be encoded in deep neural networks?

How can active learning be used in the case of complex models that require a lot of data?

What's going on in the equation of the variational lower bound?

How does the Kullback-Leibler divergence give "knowledge gained"?

How does size of the dataset depend on VC dimension of the hypothesis class?

What is the difference between hill-climbing and greedy best-first search algorithms?

What are the typical sizes of practical/commercial artificial neural networks?

Can Reinforcement Learning be used to generate sequences?

Does elitism cause premature convergence in genetic algorithms?

What are the main ideas behind NER?

Any comparison between transformer and RNN+Attention on the same dataset?

Is the VC Dimension meaningful in the context of Reinforcement Learning?

Which metric should I use to assess the quality of the clusters?

Is knowing underlying probability distribution mandatory for deciding iid property of random variables?

What's the threshold to call something 'machine learning'?

What does "Gau" in GauGAN stand for?

Are there deep networks that can differentiate object class from individual object?

Understanding relation between VC Symmetrization Lemma and Generalization Bounds

What kind of problems cannot be solved using machine learning techniques?

Why does estimation error increase with $|H|$ and decrease with $m$ in PAC learning?

How do I prove that $\mathcal{H}$, with $\mathcal{VC}$ dimension $d$, shatters all subsets with size less than $d-1$?

What does the notation ${s'\sim T(s,a,\cdot)}$ mean?

What is the name of an AI system that learns by trial and error?

Why does the hill climbing algorithm only produce a local maximum?

Why does Google Translate produce two different translations for the same Chinese text?

What is the difference between meta-learning and zero-shot learning?

Do I need an encoder-decoder architecture to predict the next item of a sequence?