User hanugm

5

votes

2

answers

313

views

Is an embedding a representation of a word or its meaning?

May 29 at 13:01 nbro♦ 30.1k

1

4

votes

2

answers

412

views

When can I call an entity a hyperparameter?

machine-learning definitions hyper-parameters

Oct 29 at 15:10 Neil Slater 21k

1

4

votes

1

answer

164

views

What is numerical stability?

neural-networks machine-learning terminology numerical-algorithms numerical-stability

Aug 3 at 14:11 OmG 1,535

4

votes

2

answers

107

views

Is there any difference between affine transformation and linear transformation?

neural-networks terminology pytorch linear-algebra transformation

Jul 19 at 10:43 y-prudent 21

4

votes

1

answer

243

views

What does "statistical efficiency" mean in this context?

convolutional-neural-networks terminology books

Apr 19 at 11:34 nbro♦ 30.1k

1

4

votes

1

answer

195

views

How can a probability density value be used for the likelihood calculation?

machine-learning comparison probability-distribution maximum-likelihood

Jan 8 at 12:53 David Ireland 3,893

3

votes

0

answers

18

views

Do any practical deep learning algorithms deal with tensors containing non-real entries?

deep-learning datasets real-world tensor

Nov 22 at 11:42 nbro♦ 30.1k

3

votes

1

answer

45

views

Can I apply reparametrization trick on "any" deep neural network?

neural-networks training autoencoders

Nov 19 at 11:36 nbro♦ 30.1k

3

votes

2

answers

106

views

In layman terms, what does "attention" do in a transformer?

neural-networks papers transformer attention

Oct 20 at 12:45 hH1sG0n3 121

3

votes

1

answer

39

views

Is there any reason behind bias towards max pooling over avg pooling?

deep-learning pooling max-pooling pooling-layers average-pooling

Oct 18 at 20:28 nbro♦ 30.1k

1

3

votes

2

answers

49

views

What are the mathematical properties of natural exponential function that lead to its usefulness in artificial intelligence?

natural-exponential-function

Oct 16 at 11:05 CommunityBot 1

1

3

votes

0

answers

77

views

What are the purposes of pooling in CNNs?

convolutional-neural-networks pooling

Sep 30 at 0:42 hanugm 2,619

3

votes

1

answer

246

views

What does "linear unit" mean in the names of activation functions?

neural-networks deep-learning terminology activation-function

Sep 17 at 16:24 nbro♦ 30.1k

3

votes

1

answer

57

views

What are the necessary mathematical properties to be a loss function in gradient based optimizations?

deep-learning objective-functions math optimization gradient-descent

Jul 28 at 11:46 nbro♦ 30.1k

3

votes

1

answer

147

views

Why is tanh a "smoothly" differentiable function?

neural-networks terminology math activation-function tanh

Jul 8 at 13:38 htl 951

3

votes

2

answers

83

views

Why do we commonly use the $\log$ to squash frequencies?

natural-language-processing definitions books tf-idf logarithm

Jun 16 at 16:06 Aray Karjauv 617

3

votes

1

answer

122

views

Where do the feature extraction and representation learning differ?

machine-learning comparison terminology feature-extraction representation-learning

May 28 at 13:35 nbro♦ 30.1k

3

votes

1

answer

165

views

How is the state-value function expressed as a product of sums?

reinforcement-learning math value-functions books expectation

May 10 at 13:45 nbro♦ 30.1k

3

votes

4

answers

649

views

What is the fundamental difference between an ML model and a function?

machine-learning comparison math definitions models

Mar 2 at 10:18 nbro♦ 30.1k

3

votes

1

answer

998

views

What is the definition of the hinge loss function?

neural-networks definitions support-vector-machine binary-classification hinge-loss

Feb 11 at 13:26 nbro♦ 30.1k

1

3

votes

1

answer

130

views

How do AI researchers imagine higher dimensions?

research academia dimensionality-reduction data-visualization

Oct 13 '20 at 22:40 nbro♦ 30.1k

2

votes

1

answer

29

views

What is the stride information of an image referring here?

deep-learning convolutional-neural-networks terminology image-processing stride

2d ago technik 1

2

votes

1

answer

36

views

Is there any simple example for volumetric data except from physics and medicine?

datasets real-world

Nov 21 at 15:51 Avatrin 366

2

votes

1

answer

19

views

Are the domains of objective functions in AI always equals to $\mathbb{R}^D$ or subset of it?

math objective-functions

Nov 6 at 5:28 spiridon_the_sun_rotator 1,389

2

votes

1

answer

42

views

What is the analytical formula for "Kaiming He" probability density function?

convolutional-neural-networks math probability-distribution weights-initialization

Nov 5 at 1:39 hanugm 2,619

1

2

votes

1

answer

136

views

What exactly happens in gradient clipping by norm?

deep-learning pytorch gradient-clipping

Nov 1 at 11:14 Archana David 301

2

votes

2

answers

90

views

Why does critical points and stationary points are used interchangeably?

terminology books derivative

Oct 25 at 0:30 hanugm 2,619

2

votes

1

answer

91

views

Do researchers generally treat tensors just as mathematical objects with certain shape?

research data-visualization tensor

Oct 21 at 1:39 hanugm 2,619

2

votes

2

answers

47

views

When would it make sense to perform a gradient descent step for each term of a loss function with multiple terms?

deep-learning objective-functions mini-batch-gradient-descent

Oct 15 at 7:08 serali 917

2

votes

0

answers

40

views

Is stability an attribute of model or training algorithm used or combination of both?

neural-networks training stability

Oct 9 at 8:31 hanugm 2,619

Stack Exchange Network

current community

your communities

more stack exchange communities

Next privilege

Next badge

232 Questions

Is an embedding a representation of a word or its meaning?

When can I call an entity a hyperparameter?

What is numerical stability?

Is there any difference between affine transformation and linear transformation?

What does "statistical efficiency" mean in this context?

How can a probability density value be used for the likelihood calculation?

Do any practical deep learning algorithms deal with tensors containing non-real entries?

Can I apply reparametrization trick on "any" deep neural network?

In layman terms, what does "attention" do in a transformer?

Is there any reason behind bias towards max pooling over avg pooling?

What are the mathematical properties of natural exponential function that lead to its usefulness in artificial intelligence?

What are the purposes of pooling in CNNs?

What does "linear unit" mean in the names of activation functions?

What are the necessary mathematical properties to be a loss function in gradient based optimizations?

Why is tanh a "smoothly" differentiable function?

Why do we commonly use the $\log$ to squash frequencies?

Where do the feature extraction and representation learning differ?

How is the state-value function expressed as a product of sums?

What is the fundamental difference between an ML model and a function?

What is the definition of the hinge loss function?

How do AI researchers imagine higher dimensions?

What is the stride information of an image referring here?

Is there any simple example for volumetric data except from physics and medicine?

Are the domains of objective functions in AI always equals to $\mathbb{R}^D$ or subset of it?

What is the analytical formula for "Kaiming He" probability density function?

What exactly happens in gradient clipping by norm?

Why does critical points and stationary points are used interchangeably?

Do researchers generally treat tensors just as mathematical objects with certain shape?

When would it make sense to perform a gradient descent step for each term of a loss function with multiple terms?

Is stability an attribute of model or training algorithm used or combination of both?