question_id stringlengths 24 37 | question stringlengths 150 27.3k | answer_id stringlengths 24 38 | answer stringlengths 33 18.3k | vote_score int32 -4 306 | accepted bool 2
classes | disagreement_type stringclasses 3
values | updated_at stringdate 2014-05-14 17:06:33 2026-08-03 11:04:45 | author stringlengths 3 29 | source_url stringlengths 32 46 | license stringclasses 2
values |
|---|---|---|---|---|---|---|---|---|---|---|
ai.stackexchange.com:q:1 | [TITLE]
What is "backprop"?
[QUESTION]
What does "backprop" mean? Is the "backprop" term basically the same as "backpropagation" or does it have a different meaning?
[TAGS]
backpropagation, definitions, neural-networks, terminology | ai.stackexchange.com:a:3 | "Backprop" is the same as "backpropagation": it's just a shorter way to say it. It is sometimes abbreviated as "BP". | 15 | true | accepted_unique_top | 2016-08-02T15:40:24Z | Franck Dernoncourt | https://ai.stackexchange.com/a/3 | CC BY-SA 3.0 |
ai.stackexchange.com:q:1 | [TITLE]
What is "backprop"?
[QUESTION]
What does "backprop" mean? Is the "backprop" term basically the same as "backpropagation" or does it have a different meaning?
[TAGS]
backpropagation, definitions, neural-networks, terminology | ai.stackexchange.com:a:83 | Yes, as Franck has rightly put, "backprop" means backpropogation, which is frequently used in the domain of neural networks for error optimization.
For a detailed explanation, I would point out
this tutorial
on the concept of backpropogation by a very good book of Michael Nielsen. | 3 | false | accepted_unique_top | 2016-08-02T16:54:40Z | Dawny33 | https://ai.stackexchange.com/a/83 | CC BY-SA 3.0 |
ai.stackexchange.com:q:1 | [TITLE]
What is "backprop"?
[QUESTION]
What does "backprop" mean? Is the "backprop" term basically the same as "backpropagation" or does it have a different meaning?
[TAGS]
backpropagation, definitions, neural-networks, terminology | ai.stackexchange.com:a:222 | 'Backprop' is short for 'backpropagation of error' in order to avoid confusion when using
backpropagation
term.
Basically
backpropagation
refers to the method for computing the gradient of the case-wise error function with respect to the weights for a feedforward network
Werbos
. And
backprop
refers to a training meth... | 10 | false | accepted_unique_top | 2016-08-03T14:39:02Z | kenorb | https://ai.stackexchange.com/a/222 | CC BY-SA 3.0 |
ai.stackexchange.com:q:1 | [TITLE]
What is "backprop"?
[QUESTION]
What does "backprop" mean? Is the "backprop" term basically the same as "backpropagation" or does it have a different meaning?
[TAGS]
backpropagation, definitions, neural-networks, terminology | ai.stackexchange.com:a:20534 | It's a fancy name for the multivariable chain rule. | 1 | false | accepted_unique_top | 2020-04-22T00:49:46Z | FourierFlux | https://ai.stackexchange.com/a/20534 | CC BY-SA 4.0 |
ai.stackexchange.com:q:1 | [TITLE]
What is "backprop"?
[QUESTION]
What does "backprop" mean? Is the "backprop" term basically the same as "backpropagation" or does it have a different meaning?
[TAGS]
backpropagation, definitions, neural-networks, terminology | ai.stackexchange.com:a:28584 | We need to compute the gradients in-order to train the deep neural networks. Deep neural network consists of many layers. Weight parameters are present between the layers. Since we need to compute the gradients of loss function for each weight, we use an algorithm called backprop. It is an abbreviation for
backprop
aga... | 0 | false | accepted_unique_top | 2021-07-08T10:45:23Z | hanugm | https://ai.stackexchange.com/a/28584 | CC BY-SA 4.0 |
ai.stackexchange.com:q:10013 | [TITLE]
What is a bad local minimum in machine learning?
[QUESTION]
What is "bad local minima"?
The following papers all mention this expression.
- Eliminating all bad Local Minima from Loss Landscapes without even adding an Extra Unit
- limination of All Bad Local Minima in Deep Learning
- Adding One Neuron Can El... | ai.stackexchange.com:a:10017 | As mentioned in the abstract of on of these papers, bad local minima is a suboptimal local minimum which means a local minimum that is near to a global minimum. | 0 | false | accepted_unique_top | 2019-01-16T07:46:43Z | OmG | https://ai.stackexchange.com/a/10017 | CC BY-SA 4.0 |
ai.stackexchange.com:q:10013 | [TITLE]
What is a bad local minimum in machine learning?
[QUESTION]
What is "bad local minima"?
The following papers all mention this expression.
- Eliminating all bad Local Minima from Loss Landscapes without even adding an Extra Unit
- limination of All Bad Local Minima in Deep Learning
- Adding One Neuron Can El... | ai.stackexchange.com:a:10088 | The adjective
bad
isn't mathematically descriptive. A better term is sub-optimal, which implies the state of learning might appear optimal based on current information but the optimal solution from among all possibilities is not yet located.
Consider a graph representing a loss function, one of the names to measure di... | 2 | true | accepted_unique_top | 2019-01-20T14:49:32Z | Douglas Daseeco | https://ai.stackexchange.com/a/10088 | CC BY-SA 4.0 |
ai.stackexchange.com:q:10049 | [TITLE]
Why are lambda returns so rarely used in policy gradients?
[QUESTION]
I've seen the Monte Carlo return
$G_{t}$
being used in REINFORCE and the TD(
$0$
) target
$r_t + \gamma Q(s', a')$
in vanilla actor-critic. However, I've never seen someone use the lambda return
$G^{\lambda}_{t}$
in these situations, nor in a... | ai.stackexchange.com:a:10061 | That can be done. For example, Chapter 13 of the 2nd edition of Sutton and Barto's Reinforcement Learning book (page 332) has pseudocode for "Actor Critic with Eligibility Traces". It's using
$G_t^{\lambda}$
returns for the critic (value function estimator), but also for the actor's policy gradients.
Note that you do ... | 11 | true | accepted_unique_top | 2019-01-18T14:29:37Z | Dennis Soemers | https://ai.stackexchange.com/a/10061 | CC BY-SA 4.0 |
ai.stackexchange.com:q:10049 | [TITLE]
Why are lambda returns so rarely used in policy gradients?
[QUESTION]
I've seen the Monte Carlo return
$G_{t}$
being used in REINFORCE and the TD(
$0$
) target
$r_t + \gamma Q(s', a')$
in vanilla actor-critic. However, I've never seen someone use the lambda return
$G^{\lambda}_{t}$
in these situations, nor in a... | ai.stackexchange.com:a:17095 | Recent actor-critic algorithms
do
use
$\lambda$
-returns, but they are disguised as something called the
Generalized Advantage Estimator
defined as
$A^{GAE}_t = \sum_{i=0}^{\infty} (\gamma\lambda)^i \delta_{t+i}$
where
$\delta_t = r_t + \gamma V(s_{t+1}) - V(s_t)$
. This turns out to be identically equal to
$[G^\lambda... | 9 | false | accepted_unique_top | 2020-08-11T16:59:36Z | Brett Daley | https://ai.stackexchange.com/a/17095 | CC BY-SA 4.0 |
ai.stackexchange.com:q:10114 | [TITLE]
What's the commercial usage of "image captioning"?
[QUESTION]
If "image captioning" is utilized to make a commercial product, what application fields will need this technique? And what is the level of required performance for this technique to be usable?
[TAGS]
deep-learning, image-recognition, natural-language... | ai.stackexchange.com:a:10116 | If "image captioning" is utilized to make a commercial product, what application fields will need this technique?
There are several important use case categories for image captioning, but most are components in larger systems, web traffic control strategies, SaaS, IaaS, IoT, and virtual reality systems, not as much fo... | 0 | true | accepted_tied_top | 2020-06-17T09:57:20Z | Douglas Daseeco | https://ai.stackexchange.com/a/10116 | CC BY-SA 4.0 |
ai.stackexchange.com:q:10114 | [TITLE]
What's the commercial usage of "image captioning"?
[QUESTION]
If "image captioning" is utilized to make a commercial product, what application fields will need this technique? And what is the level of required performance for this technique to be usable?
[TAGS]
deep-learning, image-recognition, natural-language... | ai.stackexchange.com:a:49035 | A few more use cases:
- Accessibility: Providing automated alt-text for images, making websites and apps more inclusive for users with visual impairments. (which is why I'm against that
proposal
on MSE of making image description (a.k.a. alt texts) required, since AI could do it).
- Author assistance: Automatically c... | 0 | false | accepted_tied_top | 2025-10-13T01:37:33Z | Franck Dernoncourt | https://ai.stackexchange.com/a/49035 | CC BY-SA 4.0 |
ai.stackexchange.com:q:10133 | [TITLE]
What are the segment embeddings and position embeddings in BERT?
[QUESTION]
They only reference in the paper that the position embeddings are learned, which is different from what was done in ELMo.
ELMo paper -
https://arxiv.org/pdf/1802.05365.pdf
BERT paper -
https://arxiv.org/pdf/1810.04805.pdf
[TAGS]
bert,... | ai.stackexchange.com:a:10630 | These embeddings are nothing more than token embeddings.
You just randomly initialize them, then use gradient descent to train them, just like what you do with token embeddings. | 3 | true | accepted_below_top | 2019-02-17T09:16:51Z | soloice | https://ai.stackexchange.com/a/10630 | CC BY-SA 4.0 |
ai.stackexchange.com:q:10133 | [TITLE]
What are the segment embeddings and position embeddings in BERT?
[QUESTION]
They only reference in the paper that the position embeddings are learned, which is different from what was done in ELMo.
ELMo paper -
https://arxiv.org/pdf/1802.05365.pdf
BERT paper -
https://arxiv.org/pdf/1810.04805.pdf
[TAGS]
bert,... | ai.stackexchange.com:a:18072 | Sentences (for those tasks such as NLI which take two sentences as input) are differentiated in two ways in BERT:
- First, a
`[SEP]`
token is put between them
- Second, a learned embedding
$E_A$
is added to every token of the first sentence, and another learned vector
$E_B$
to every token of the second one
That is, ... | 4 | false | accepted_below_top | 2023-10-09T17:31:40Z | finiteautomata | https://ai.stackexchange.com/a/18072 | CC BY-SA 4.0 |
ai.stackexchange.com:q:10133 | [TITLE]
What are the segment embeddings and position embeddings in BERT?
[QUESTION]
They only reference in the paper that the position embeddings are learned, which is different from what was done in ELMo.
ELMo paper -
https://arxiv.org/pdf/1802.05365.pdf
BERT paper -
https://arxiv.org/pdf/1810.04805.pdf
[TAGS]
bert,... | ai.stackexchange.com:a:49048 | Segment embedding are used to distinguish two separate splits of a sentence:
Example - My dog is cute, he like playing
Will be split into: [My dog is cute] [SEP] [he likes play ing]
source | 0 | false | accepted_below_top | 2025-10-14T19:50:36Z | Loony_D | https://ai.stackexchange.com/a/49048 | CC BY-SA 4.0 |
ai.stackexchange.com:q:10158 | [TITLE]
How can a neural network learn when the derivative of the activation function is 0?
[QUESTION]
Imagine that I have an artificial neural network with a single hidden layer and that I am using ReLU as my activating function.
If by change I initialize my bias and my weights in such a form that:
$$
X * W + B < 0
$... | ai.stackexchange.com:a:10160 | In a setup like the above where the derivat[iv]e is 0 is it true that an NN won´t learn anything?
There are a couple of adjustments to gradients that
might
apply if you do this in a standard framework:
- Momentum may cause weights to continue changing if any recent ones were non-zero. This is typically implemented as... | 3 | true | accepted_unique_top | 2019-01-24T08:21:49Z | Neil Slater | https://ai.stackexchange.com/a/10160 | CC BY-SA 4.0 |
ai.stackexchange.com:q:10158 | [TITLE]
How can a neural network learn when the derivative of the activation function is 0?
[QUESTION]
Imagine that I have an artificial neural network with a single hidden layer and that I am using ReLU as my activating function.
If by change I initialize my bias and my weights in such a form that:
$$
X * W + B < 0
$... | ai.stackexchange.com:a:10165 | Learning and Zero Derivatives
Artificial networks are designed so that even when the partial derivative of a single activation function is zero they can learn. They can also be designed to continue learning when the derivative of the loss
1
function is zero too. This resilience to a vanishing feedback signal amplitude... | 1 | false | accepted_unique_top | 2019-01-28T17:13:29Z | Douglas Daseeco | https://ai.stackexchange.com/a/10165 | CC BY-SA 4.0 |
ai.stackexchange.com:q:10158 | [TITLE]
How can a neural network learn when the derivative of the activation function is 0?
[QUESTION]
Imagine that I have an artificial neural network with a single hidden layer and that I am using ReLU as my activating function.
If by change I initialize my bias and my weights in such a form that:
$$
X * W + B < 0
$... | ai.stackexchange.com:a:10185 | People often place a batchnorm layer before ReLU. That effectively prevents the problem you have described. | 1 | false | accepted_unique_top | 2019-01-25T12:31:58Z | ssegvic | https://ai.stackexchange.com/a/10185 | CC BY-SA 4.0 |
ai.stackexchange.com:q:10177 | [TITLE]
What's the role of bounding boxes in object detection?
[QUESTION]
I'm quite new to the field of computer vision and was wondering what are the purposes of having the boundary boxes in object detection.
Obviously, it shows where the detected object is, and using a classifier can only classify one object per ima... | ai.stackexchange.com:a:10186 | In principle, you could train the model to output a sigmoid map of coarse object positions (0 -> no object, 1 -> an object center is located here). The map could be subjected to non-maximum suppression and such model could be trained end-to-end. That would be possible, if that's what you are asking. | -1 | false | accepted_unique_top | 2019-01-25T12:52:43Z | ssegvic | https://ai.stackexchange.com/a/10186 | CC BY-SA 4.0 |
ai.stackexchange.com:q:10177 | [TITLE]
What's the role of bounding boxes in object detection?
[QUESTION]
I'm quite new to the field of computer vision and was wondering what are the purposes of having the boundary boxes in object detection.
Obviously, it shows where the detected object is, and using a classifier can only classify one object per ima... | ai.stackexchange.com:a:10221 | A bounding box is a rectangle superimposed over an image within which all important features of a particular object is expected to reside. It's purpose is to reduce the range of search for those object features and thereby conserve computing resources: Allocation of memory, processors, cores, processing time, some othe... | 3 | true | accepted_unique_top | 2019-01-28T11:16:27Z | Douglas Daseeco | https://ai.stackexchange.com/a/10221 | CC BY-SA 4.0 |
ai.stackexchange.com:q:10289 | [TITLE]
Are neural networks statistical models?
[QUESTION]
By reading the abstract of
Neural Networks and Statistical Models
paper it would seem that ANNs are statistical models.
In contrast
Machine Learning is not just glorified Statistics
.
I am looking for a more concise/summarized answer with focus on ANNs.
[TAGS... | ai.stackexchange.com:a:12354 | What is a statistical model?
According to Anthony C. Davison (in the book
Statistical Models
), a statistical model is a probability distribution constructed to enable inferences to be drawn or decisions made from data. The probability distribution represents the variability of the data.
Are neural networks statistic... | 7 | true | accepted_unique_top | 2022-06-04T14:15:06Z | nbro | https://ai.stackexchange.com/a/12354 | CC BY-SA 4.0 |
ai.stackexchange.com:q:10289 | [TITLE]
Are neural networks statistical models?
[QUESTION]
By reading the abstract of
Neural Networks and Statistical Models
paper it would seem that ANNs are statistical models.
In contrast
Machine Learning is not just glorified Statistics
.
I am looking for a more concise/summarized answer with focus on ANNs.
[TAGS... | ai.stackexchange.com:a:18580 | According to Wikipedia:
A statistical model is a mathematical model that embodies a set of statistical assumptions concerning the generation of sample data (and similar data from a larger population). A statistical model represents, often in considerably idealized form, the data-generating process.
Answer to your que... | 5 | false | accepted_unique_top | 2020-03-12T03:05:38Z | Ta_Req | https://ai.stackexchange.com/a/18580 | CC BY-SA 4.0 |
ai.stackexchange.com:q:10289 | [TITLE]
Are neural networks statistical models?
[QUESTION]
By reading the abstract of
Neural Networks and Statistical Models
paper it would seem that ANNs are statistical models.
In contrast
Machine Learning is not just glorified Statistics
.
I am looking for a more concise/summarized answer with focus on ANNs.
[TAGS... | ai.stackexchange.com:a:37638 | A dataset can be thought of as a set of ordered pairs
$\subset R^f \times R^l$
, where
$f$
is the feature dimensions and
$l$
the label dimensions.
Ordered pairs give rise to a statistical model (i.e. a function from
$R^f$
to a probability distribution over
$R^l$
)
Then this statistical model is turned into function
$... | 2 | false | accepted_unique_top | 2022-10-26T05:17:54Z | Tom Huntington | https://ai.stackexchange.com/a/37638 | CC BY-SA 4.0 |
ai.stackexchange.com:q:10306 | [TITLE]
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?
[QUESTION]
Status
:
For a few weeks now, I have been working on a Double DQN agent for the
`PongDeterministic-v4`
environment, which you can find
here
.
A single ... | ai.stackexchange.com:a:10311 | Dueling architectures create bigger differences in the values of actions in the state space. This is because the state-value
V(s)
function is estimated separately from the state-action value
Q(s, a)
. A new quantity, the advantage of an action, can then be defined as
A(s, a) = Q(s, a) - V(s)
.
The Q function, however,... | 3 | false | accepted_tied_top | 2019-01-30T15:17:46Z | Jaden Travnik | https://ai.stackexchange.com/a/10311 | CC BY-SA 4.0 |
ai.stackexchange.com:q:10306 | [TITLE]
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?
[QUESTION]
Status
:
For a few weeks now, I have been working on a Double DQN agent for the
`PongDeterministic-v4`
environment, which you can find
here
.
A single ... | ai.stackexchange.com:a:10481 | Although what @Jaden said may be true by itself, it does not really serve to answer my question as I have seen after conducting numerous experiments, and finally reaching close to Dueling Network performance using a normal Double DQN (DDQN).
I made the following changes to my code after closely examining the OpenAI ba... | 3 | true | accepted_tied_top | 2019-02-10T03:56:15Z | hridayns | https://ai.stackexchange.com/a/10481 | CC BY-SA 4.0 |
ai.stackexchange.com:q:10322 | [TITLE]
Using GAN's to generate dataset for CNN training
[QUESTION]
I'm doing bachaleor thesis on traffic sign detection using single shot detector called YOLO. These single shot detectors can perform detection of objects in image and so they have specific way of training, ie. training on full images. Thats quite probl... | ai.stackexchange.com:a:10325 | You could add the desired traffic sign location to the latent vector and then arrange that the generator incurs loss if the traffic sign is not at the right place in the generated image. | 0 | false | accepted_unique_top | 2019-01-31T11:57:59Z | ssegvic | https://ai.stackexchange.com/a/10325 | CC BY-SA 4.0 |
ai.stackexchange.com:q:10322 | [TITLE]
Using GAN's to generate dataset for CNN training
[QUESTION]
I'm doing bachaleor thesis on traffic sign detection using single shot detector called YOLO. These single shot detectors can perform detection of objects in image and so they have specific way of training, ie. training on full images. Thats quite probl... | ai.stackexchange.com:a:10328 | I think you'll
enjoy this work from Apple
on improving the realism of synthetic images. Essentially what you need to do is generate a synthetic image
then
have your GAN
modify
the synthetic image so that a 1) a discriminator thinks it is real while also 2) not changing the gross structure of the image very much (so the... | 3 | true | accepted_unique_top | 2019-01-31T13:03:24Z | Edward Dixon | https://ai.stackexchange.com/a/10328 | CC BY-SA 4.0 |
ai.stackexchange.com:q:10352 | [TITLE]
When are Q values calculated in experience replay?
[QUESTION]
In
experience replay
, the update rule follows the loss:
$$
L_i(\theta_i) = \mathbb{E}_{(s_t, a_t, r_t, s_{t+1}) \sim U(D)} \left[ \left(r_t + \gamma \max_{a_{t+1}} Q(s_{t+1}, a_{t+1}; \theta_i^-) - Q(s_t, a_t; \theta_i)\right)^2 \right]
$$
I can't... | ai.stackexchange.com:a:10354 | Once the algorithm reaches a point where a reward (or penalty) is gained, then a slowly decaying amount of that reward is given to the Q-value of whichever action was derived from a particular state for each time step back through the game (or however far back you want to apply that reward).
You will have stored the st... | 0 | false | accepted_unique_top | 2019-02-01T19:11:22Z | DrMcCleod | https://ai.stackexchange.com/a/10354 | CC BY-SA 4.0 |
ai.stackexchange.com:q:10352 | [TITLE]
When are Q values calculated in experience replay?
[QUESTION]
In
experience replay
, the update rule follows the loss:
$$
L_i(\theta_i) = \mathbb{E}_{(s_t, a_t, r_t, s_{t+1}) \sim U(D)} \left[ \left(r_t + \gamma \max_{a_{t+1}} Q(s_{t+1}, a_{t+1}; \theta_i^-) - Q(s_t, a_t; \theta_i)\right)^2 \right]
$$
I can't... | ai.stackexchange.com:a:10362 | Because the Q value is different, I don't see how the reward signal at time
$t$
is of any relevance for
$Q_{t+x}(s_t,a_t)$
at
$t+x$
, the time of learning.
The
$r_t$
value for any single step is not dependent on
$Q$
or the current policy. It is purely dependent on
$(s_t,a_t)$
. That means you can use the Q update equa... | 2 | true | accepted_unique_top | 2019-02-02T09:45:29Z | Neil Slater | https://ai.stackexchange.com/a/10362 | CC BY-SA 4.0 |
ai.stackexchange.com:q:10431 | [TITLE]
Should I choose a model with the smallest loss or highest accuracy?
[QUESTION]
I have two Machine Learning models (I use LSTM) that have a different result on the validation set (~100 samples data):
- Model A: Accuracy: ~91%, Loss: ~0.01
- Model B: Accuracy: ~83%, Loss: ~0.003
The size and the speed of both ... | ai.stackexchange.com:a:10588 | You should choose the model A. The loss is just a differentiable proxy for accuracy.
That said, the situation should be examined in more detail. If the higher loss is due to the data term, examine the data which produce high loss and check for presence of overfitting or incorrect labels.
If the higher loss is due to ... | 7 | false | accepted_below_top | 2019-02-15T18:17:09Z | ssegvic | https://ai.stackexchange.com/a/10588 | CC BY-SA 4.0 |
ai.stackexchange.com:q:10431 | [TITLE]
Should I choose a model with the smallest loss or highest accuracy?
[QUESTION]
I have two Machine Learning models (I use LSTM) that have a different result on the validation set (~100 samples data):
- Model A: Accuracy: ~91%, Loss: ~0.01
- Model B: Accuracy: ~83%, Loss: ~0.003
The size and the speed of both ... | ai.stackexchange.com:a:10589 | It depends on your application! Imagine a binary classifier that is always very "confident" - it always assigns P=100% to Class A and 0% to Class B, or
vice versa
(sometimes wrong, never uncertain!). Now imagine a "humble" model that is perhaps fractionally less accurate, but whose probabilities are actually meaningful... | 1 | false | accepted_below_top | 2019-02-15T11:07:42Z | Edward Dixon | https://ai.stackexchange.com/a/10589 | CC BY-SA 4.0 |
ai.stackexchange.com:q:10431 | [TITLE]
Should I choose a model with the smallest loss or highest accuracy?
[QUESTION]
I have two Machine Learning models (I use LSTM) that have a different result on the validation set (~100 samples data):
- Model A: Accuracy: ~91%, Loss: ~0.01
- Model B: Accuracy: ~83%, Loss: ~0.003
The size and the speed of both ... | ai.stackexchange.com:a:10604 | You should note that both your results are consistent with a "true" probability of 87% accuracy, and your measurement of a difference between these models is not statistically significant. With an 87% accuracy applied at random, then there is approx 14% chance of getting the two extremes of accuracy you have observed b... | 5 | true | accepted_below_top | 2019-02-15T19:40:51Z | Neil Slater | https://ai.stackexchange.com/a/10604 | CC BY-SA 4.0 |
ai.stackexchange.com:q:10447 | [TITLE]
How to use CNN for making predictions on non-image data?
[QUESTION]
I have a dataset which I have loaded as a data frame in Python. It consists of 21392 rows (the data instances, each row is one sample) and 1972 columns (the features). The last column i.e. column 1972 has string type labels (14 different catego... | ai.stackexchange.com:a:10449 | You can use CNN on any data, but it's
recommended
to use CNN only on data that have spatial features (It might still work on data that doesn't have spatial features, see DuttaA's comment below).
For example, in the image, the connection between pixels in some area gives you another feature (e.g. edge) instead of a fea... | 12 | true | accepted_unique_top | 2021-04-13T15:54:51Z | malioboro | https://ai.stackexchange.com/a/10449 | CC BY-SA 4.0 |
ai.stackexchange.com:q:10447 | [TITLE]
How to use CNN for making predictions on non-image data?
[QUESTION]
I have a dataset which I have loaded as a data frame in Python. It consists of 21392 rows (the data instances, each row is one sample) and 1972 columns (the features). The last column i.e. column 1972 has string type labels (14 different catego... | ai.stackexchange.com:a:10458 | The convolutional models are a method of choice when your problem is translation invariant (or covariant). In image classification, the image should be classified into class 'cow' if a cow is present in any part of the image. In text classification, different orders of phrases and sentences result in related meaning. I... | 6 | false | accepted_unique_top | 2019-02-08T16:05:53Z | ssegvic | https://ai.stackexchange.com/a/10458 | CC BY-SA 4.0 |
ai.stackexchange.com:q:10447 | [TITLE]
How to use CNN for making predictions on non-image data?
[QUESTION]
I have a dataset which I have loaded as a data frame in Python. It consists of 21392 rows (the data instances, each row is one sample) and 1972 columns (the features). The last column i.e. column 1972 has string type labels (14 different catego... | ai.stackexchange.com:a:30220 | DeepInsight method has been used for converting tabular data into corresponding images which are then processed by CNN. Here is the link
https://alok-ai-lab.github.io/DeepInsight/ | 1 | false | accepted_unique_top | 2021-08-17T08:34:56Z | Astha | https://ai.stackexchange.com/a/30220 | CC BY-SA 4.0 |
ai.stackexchange.com:q:10529 | [TITLE]
How to obtain a formula for loss, when given an iterative update rule in gradient descent?
[QUESTION]
From
the reinforcement learning book section 13.3
:
Using pytorch, I need to calculate a loss, and then the gradient is calculated internally.
How to obtain the loss from equations which are stated in the for... | ai.stackexchange.com:a:10538 | You can find an implementation of the REINFORCE algorithm (as defined in your question) in PyTorch at the following URL:
https://github.com/JamesChuanggg/pytorch-REINFORCE/
. First of all, I would like to note that a policy can be represented or implemented as a neural network, where the input is the state (you are cur... | 3 | true | accepted_unique_top | 2019-02-12T18:09:22Z | nbro | https://ai.stackexchange.com/a/10538 | CC BY-SA 4.0 |
ai.stackexchange.com:q:10529 | [TITLE]
How to obtain a formula for loss, when given an iterative update rule in gradient descent?
[QUESTION]
From
the reinforcement learning book section 13.3
:
Using pytorch, I need to calculate a loss, and then the gradient is calculated internally.
How to obtain the loss from equations which are stated in the for... | ai.stackexchange.com:a:27423 | Chiming in because I had the same question and stumbled across your post. It seems like the general version of your question still has not been answered.
In general, a well-formed gradient update rule is all you need to be able to train the network. We are thinking of converting to a "loss function" because that is th... | 0 | false | accepted_unique_top | 2021-04-20T01:51:16Z | killian95 | https://ai.stackexchange.com/a/27423 | CC BY-SA 4.0 |
ai.stackexchange.com:q:10549 | [TITLE]
What is the gradient of the objective function in the Soft Actor-Critic paper?
[QUESTION]
In the paper
Soft Actor-Critic: Off-Policy Maximum Entropy Deep Reinforcement Learning with a Stochastic Actor
, they define the loss function for the policy network as
$$
J_\pi(\phi)=\mathbb E_{s_t\sim \mathcal D}\left[D... | ai.stackexchange.com:a:10573 | I'll give it a go here and try to answer your question, I'm not sure if this is entirely correct, so if someone thinks that it isn't please correct me.
I'll disregard expectation here to make things simpler. First, note that policy
$\pi$
depends on parameter vector
$\phi$
and function
$f_\phi(\epsilon_t;s_t)$
, and va... | 4 | true | accepted_unique_top | 2019-02-14T15:30:03Z | Brale | https://ai.stackexchange.com/a/10573 | CC BY-SA 4.0 |
ai.stackexchange.com:q:10549 | [TITLE]
What is the gradient of the objective function in the Soft Actor-Critic paper?
[QUESTION]
In the paper
Soft Actor-Critic: Off-Policy Maximum Entropy Deep Reinforcement Learning with a Stochastic Actor
, they define the loss function for the policy network as
$$
J_\pi(\phi)=\mathbb E_{s_t\sim \mathcal D}\left[D... | ai.stackexchange.com:a:23846 | This is more meant like a comment to the previous answer. I also originally thought that
$$ \nabla_{\theta}\log \pi_{\theta}(f_{\theta}(\varepsilon, s)\mid s) = \nabla_{a}\log\pi_{\theta}(a\mid s)\vert_{a=f_{\theta}(\varepsilon,s)}\nabla_{\theta}f_{\theta}(\varepsilon, s), $$
instead of
$$ \nabla_{\theta}\log \pi_{\t... | 1 | false | accepted_unique_top | 2020-09-30T20:28:50Z | matorbi | https://ai.stackexchange.com/a/23846 | CC BY-SA 4.0 |
ai.stackexchange.com:q:10575 | [TITLE]
Apart from the state and state-action value functions, what are other examples of value functions used in RL?
[QUESTION]
In reinforcement learning, we often define two functions, the
state-value function
$$V^\pi(s) = \mathbb{E}_{\pi} \left[\sum_{k=0}^{\infty}
\gamma^{k}R_{t+k+1} \Bigg| S_t=s \right]$$
and the... | ai.stackexchange.com:a:11069 | Advantage function
:
$A(s,a) = Q(s,a) - V(s)$
More interesting is the
General Value Function (GVF)
, the expected sum of the (discounted) future values of some arbitrary signal, not necessarily reward. It is therefore a generalization of value function
$V(s)$
. The GVF is defined on page 459 of the 2nd edition of
Sutt... | 6 | true | accepted_unique_top | 2020-11-23T14:03:51Z | Philip Raeisghasem | https://ai.stackexchange.com/a/11069 | CC BY-SA 4.0 |
ai.stackexchange.com:q:10575 | [TITLE]
Apart from the state and state-action value functions, what are other examples of value functions used in RL?
[QUESTION]
In reinforcement learning, we often define two functions, the
state-value function
$$V^\pi(s) = \mathbb{E}_{\pi} \left[\sum_{k=0}^{\infty}
\gamma^{k}R_{t+k+1} \Bigg| S_t=s \right]$$
and the... | ai.stackexchange.com:a:24292 | There is also the simpler
action value function
$$q_*(a) = \mathbb{E} \left[ R_t
\mid A_t = a\right],$$
which we try to approximate when solving
context-free
bandit problems
. You can also similarly define the action value function for
contextual
bandit problems by also conditioning on the context (rather than just o... | 0 | false | accepted_unique_top | 2020-11-23T14:13:36Z | nbro | https://ai.stackexchange.com/a/24292 | CC BY-SA 4.0 |
ai.stackexchange.com:q:10615 | [TITLE]
What is "planning" in the context of reinforcement learning, and how is it different from RL and SL?
[QUESTION]
This is an excerpt taken from
Sutton and Barto
(pg. 3):
Another key feature of reinforcement learning is that it explicitly considers the whole problem of a goal-directed agent interacting with an un... | ai.stackexchange.com:a:10616 | The concept of "planning" is not just related to RL. In general (as the name suggests), planning consists in creating a "plan" which you will use to reach a "goal". The goal depends on the context or problem. For example, in robotics, you can use a "planning algorithm" (e.g. Dijkstra's algorithm) in order to find the p... | 14 | true | accepted_unique_top | 2020-11-21T13:15:24Z | nbro | https://ai.stackexchange.com/a/10616 | CC BY-SA 4.0 |
ai.stackexchange.com:q:10615 | [TITLE]
What is "planning" in the context of reinforcement learning, and how is it different from RL and SL?
[QUESTION]
This is an excerpt taken from
Sutton and Barto
(pg. 3):
Another key feature of reinforcement learning is that it explicitly considers the whole problem of a goal-directed agent interacting with an un... | ai.stackexchange.com:a:10618 | The
automated planning
is:
Automated planning and scheduling, sometimes denoted as simply AI Planning,
1
is a branch of artificial intelligence that concerns the realization of strategies or action sequences, typically for execution by intelligent agents, autonomous robots and unmanned vehicles. Unlike classical contr... | 0 | false | accepted_unique_top | 2019-02-16T17:02:52Z | OmG | https://ai.stackexchange.com/a/10618 | CC BY-SA 4.0 |
ai.stackexchange.com:q:10617 | [TITLE]
Do I need an encoder-decoder architecture to predict the next item of a sequence?
[QUESTION]
I am trying to understand how RNNs are used for sequence modelling.
On a
tutorial here
, it mentions that if you want to translate say a sentence from English to French you can use an encoder-decoder set-up as they des... | ai.stackexchange.com:a:10619 | You don't need to Encoder-Decoder here. When using seq2seq learning for text (for example, for translation) you need encoder-decoder to encode the words into the numeric vectors and decode the vectors into the words. Therefore, for your numerical case, you don't need an encoder or decoder to train the RNN. | 2 | true | accepted_unique_top | 2019-02-21T11:16:11Z | OmG | https://ai.stackexchange.com/a/10619 | CC BY-SA 4.0 |
ai.stackexchange.com:q:10617 | [TITLE]
Do I need an encoder-decoder architecture to predict the next item of a sequence?
[QUESTION]
I am trying to understand how RNNs are used for sequence modelling.
On a
tutorial here
, it mentions that if you want to translate say a sentence from English to French you can use an encoder-decoder set-up as they des... | ai.stackexchange.com:a:10779 | Why you need encoder and decoder here, this is not a use case for this. If you want to convert one sequence to another sequence then you can use encoder-decoder. You can use simple seq2seq learning for below purpose.
- Sequence
- Sequence Prediction
- Sequence Classification
- Sequence Generation Sequence to Sequen... | 1 | false | accepted_unique_top | 2019-02-21T06:56:37Z | Maheshwar Ligade | https://ai.stackexchange.com/a/10779 | CC BY-SA 4.0 |
ai.stackexchange.com:q:10620 | [TITLE]
How can we estimate the transition model and reward function?
[QUESTION]
In reinforcement learning (RL), there are model-based and model-free algorithms. In short, model-based algorithms use a transition model
$p(s' \mid s, a)$
and the reward function
$r(s, a)$
, even though they do not necessarily compute (or ... | ai.stackexchange.com:a:23647 | Given that model-based RL algorithms do not necessarily estimate or compute the transition model or reward function, in the case these are unknown, how can they be computed or estimated (so that they can be used by the model-based algorithms)?
A generally reliable approach to creating learned models from interacting w... | 1 | true | accepted_tied_top | 2022-01-24T11:29:02Z | Neil Slater | https://ai.stackexchange.com/a/23647 | CC BY-SA 4.0 |
ai.stackexchange.com:q:10620 | [TITLE]
How can we estimate the transition model and reward function?
[QUESTION]
In reinforcement learning (RL), there are model-based and model-free algorithms. In short, model-based algorithms use a transition model
$p(s' \mid s, a)$
and the reward function
$r(s, a)$
, even though they do not necessarily compute (or ... | ai.stackexchange.com:a:24511 | The original question about
both
the estimation of the
transition model
, often denoted as
$T$
, and the
reward function
, sometimes denoted as
$R$
, arose because I was thinking about the probability distribution often denoted as
$$\color{red}{p}\left(s^{\prime}, r \mid s, a\right) \doteq \operatorname{Pr}\left\{S_{t... | 1 | false | accepted_tied_top | 2020-11-09T12:22:19Z | nbro | https://ai.stackexchange.com/a/24511 | CC BY-SA 4.0 |
ai.stackexchange.com:q:10623 | [TITLE]
What is self-supervised learning in machine learning?
[QUESTION]
What is self-supervised learning in machine learning? How is it different from supervised learning?
[TAGS]
comparison, machine-learning, representation-learning, self-supervised-learning, supervised-learning | ai.stackexchange.com:a:10624 | Introduction
The term
self-supervised learning
(SSL) has been used (sometimes differently) in different contexts and fields, such as representation learning [
1
], neural networks, robotics [
2
], natural language processing, and reinforcement learning. In all cases, the basic idea is to
automatically
generate some ki... | 98 | true | accepted_unique_top | 2020-08-01T13:58:51Z | nbro | https://ai.stackexchange.com/a/10624 | CC BY-SA 4.0 |
ai.stackexchange.com:q:10623 | [TITLE]
What is self-supervised learning in machine learning?
[QUESTION]
What is self-supervised learning in machine learning? How is it different from supervised learning?
[TAGS]
comparison, machine-learning, representation-learning, self-supervised-learning, supervised-learning | ai.stackexchange.com:a:10743 | Self-supervised visual recognition is often applied to representation learning. Here we first learn features on unlabeled data (representation learning), and then learn the real model on features extracted from the labeled data. This especially makes sense when we have a lot of unlabeled data and few labeled data.
The... | 7 | false | accepted_unique_top | 2019-02-20T14:30:03Z | ssegvic | https://ai.stackexchange.com/a/10743 | CC BY-SA 4.0 |
ai.stackexchange.com:q:10623 | [TITLE]
What is self-supervised learning in machine learning?
[QUESTION]
What is self-supervised learning in machine learning? How is it different from supervised learning?
[TAGS]
comparison, machine-learning, representation-learning, self-supervised-learning, supervised-learning | ai.stackexchange.com:a:13749 | Self-supervised learning is when you use some parts of the samples as labels for a task that requires a good degree of comprehension to be solved. I'll emphasize these two key points, before giving an example:
- Labels are extracted from the sample
, so they can be generated automatically, with some very simple algori... | 20 | false | accepted_unique_top | 2019-08-02T22:06:05Z | David | https://ai.stackexchange.com/a/13749 | CC BY-SA 4.0 |
ai.stackexchange.com:q:10658 | [TITLE]
To what does the number of hidden layers in a neural network correspond?
[QUESTION]
In a neural network, the number of neurons in the hidden layer corresponds to the complexity of the model generated to map the inputs to output(s). More neurons creates a more complex function (and thus the ability to model more... | ai.stackexchange.com:a:10659 | More hidden layers will just escalate the possibilities amount the neurons, including the solutions from the previous hidden layers. (I will edit this once I am at home and provide you with a good link I found some time ago)
Meanwhile maybe this will help you
https://stats.stackexchange.com/questions/63152/what-does-t... | 1 | false | accepted_unique_top | 2019-02-18T15:56:04Z | Fleep | https://ai.stackexchange.com/a/10659 | CC BY-SA 4.0 |
ai.stackexchange.com:q:10658 | [TITLE]
To what does the number of hidden layers in a neural network correspond?
[QUESTION]
In a neural network, the number of neurons in the hidden layer corresponds to the complexity of the model generated to map the inputs to output(s). More neurons creates a more complex function (and thus the ability to model more... | ai.stackexchange.com:a:10683 | It was proven a feed-forward network with a single hidden layer containing a finite number of neurons can approximate continuous functions (see
Universal approximation theorem
).
More layers can't improve something that can already do "everything". But adding more layers reduces the number of necessary neurons, and re... | 1 | false | accepted_unique_top | 2019-02-18T22:40:24Z | Nyos | https://ai.stackexchange.com/a/10683 | CC BY-SA 4.0 |
ai.stackexchange.com:q:10658 | [TITLE]
To what does the number of hidden layers in a neural network correspond?
[QUESTION]
In a neural network, the number of neurons in the hidden layer corresponds to the complexity of the model generated to map the inputs to output(s). More neurons creates a more complex function (and thus the ability to model more... | ai.stackexchange.com:a:11357 | This is a very interesting question that you ask. I believe, this
post
and this
post
(written by me) well address your question. However, it deserves an explanation here.
1. Fully connected networks
The more layers you add, the more "nonlinear" your network becomes. For instance, in the case of
two spirals problem
, ... | 2 | true | accepted_unique_top | 2022-05-25T09:02:45Z | penkovsky | https://ai.stackexchange.com/a/11357 | CC BY-SA 4.0 |
ai.stackexchange.com:q:10773 | [TITLE]
How do neural networks weigh multiple inputs/features of different dimensionality?
[QUESTION]
I am confused about how neural networks weigh different features or inputs.
Consider this example. I have 3 features/inputs: an image, a dollar amount, and a rating. However, since one feature is an image, I need to r... | ai.stackexchange.com:a:10780 | As stated in your example, the three features are: an image, a price, a rating. Now, you want to build a model that uses all of these features and the simplest way to do is to feed them directly into the neural network, but it's inefficient and fundamentally flawed, due to the following reasons:
- In the first dense l... | 1 | true | accepted_unique_top | 2020-11-21T20:30:39Z | ssh | https://ai.stackexchange.com/a/10780 | CC BY-SA 4.0 |
ai.stackexchange.com:q:10773 | [TITLE]
How do neural networks weigh multiple inputs/features of different dimensionality?
[QUESTION]
I am confused about how neural networks weigh different features or inputs.
Consider this example. I have 3 features/inputs: an image, a dollar amount, and a rating. However, since one feature is an image, I need to r... | ai.stackexchange.com:a:10784 | The answer by ssh is correct. Your results could be further improved i) by extracting image features by a convolutional (instead of fully connected) architecture, and ii) by exploiting transfer learning.
To exploit transfer learning you i) pick some widely used model, eg. ResNet-18, ii) initialize it with ImageNet pre... | 0 | false | accepted_unique_top | 2019-02-21T10:53:30Z | ssegvic | https://ai.stackexchange.com/a/10784 | CC BY-SA 4.0 |
ai.stackexchange.com:q:10812 | [TITLE]
What is the difference between First-Visit Monte-Carlo and Every-Visit Monte-Carlo Policy Evaluation?
[QUESTION]
I came across these 2 algorithms, but I cannot understand the difference between these 2, both in terms of implementation as well as intuitionally.
So, what difference does the second point in both ... | ai.stackexchange.com:a:10818 | The first-visit and the every-visit Monte-Carlo (MC) algorithms are both used to solve the
prediction
problem
(or, also called, "evaluation problem"), that is, the problem of estimating the value function associated with a given (as input to the algorithms) fixed (that is, it does not change during the execution of the... | 27 | true | accepted_unique_top | 2019-02-22T18:57:14Z | nbro | https://ai.stackexchange.com/a/10818 | CC BY-SA 4.0 |
ai.stackexchange.com:q:10812 | [TITLE]
What is the difference between First-Visit Monte-Carlo and Every-Visit Monte-Carlo Policy Evaluation?
[QUESTION]
I came across these 2 algorithms, but I cannot understand the difference between these 2, both in terms of implementation as well as intuitionally.
So, what difference does the second point in both ... | ai.stackexchange.com:a:22219 | For anyone coming across this question and wants a very intuitive understanding of first and every visit monte-carlo, look at the answer given in the link provided here.
https://amp.reddit.com/r/reinforcementlearning/comments/9zkdjb/d_help_need_in_understanding_monte_carlo_first/
After looking at that intuition, then... | 1 | false | accepted_unique_top | 2020-06-28T02:29:06Z | Mugumya kevin | https://ai.stackexchange.com/a/22219 | CC BY-SA 4.0 |
ai.stackexchange.com:q:10904 | [TITLE]
What is the difference between DQN and AlphaGo Zero?
[QUESTION]
I have already implemented a relatively simple DQN on Pacman.
Now I would like to clearly understand the difference between a DQN and the techniques used by AlphaGo zero/AlphaZero and I couldn't find a place where the features of both approaches a... | ai.stackexchange.com:a:10905 | DQN and AlphaZero do not share much in terms of implementation.
However, they are based on the same Reinforcement Learning (RL) theoretical framework. If you understand terms like MDP, reward, return, value, policy, then these are interchangeable between DQN and AlphaZero. When it comes to implementation, and what eac... | 6 | true | accepted_unique_top | 2020-06-17T09:57:20Z | Neil Slater | https://ai.stackexchange.com/a/10905 | CC BY-SA 4.0 |
ai.stackexchange.com:q:10904 | [TITLE]
What is the difference between DQN and AlphaGo Zero?
[QUESTION]
I have already implemented a relatively simple DQN on Pacman.
Now I would like to clearly understand the difference between a DQN and the techniques used by AlphaGo zero/AlphaZero and I couldn't find a place where the features of both approaches a... | ai.stackexchange.com:a:10927 | You can actually combine AlphaZero-like approach with DQN:
A* + DQN | 1 | false | accepted_unique_top | 2019-02-28T13:04:20Z | mirror2image | https://ai.stackexchange.com/a/10927 | CC BY-SA 4.0 |
ai.stackexchange.com:q:10947 | [TITLE]
Does IBM Watson use machine learning?
[QUESTION]
I was reading an article on
Medium
and wanted to make it clear whether a bot created on IBM Watson is an intelligent one or unintelligent.
Simply put, there are 2 types of chatbots — unintelligent ones that act using predefined conversation flows (algorithms) wr... | ai.stackexchange.com:a:10952 | A quick glance at the
IBM Watson wikipedia page
reveals that it does indeed use machine learning. Watson is a complex computing system that uses a variety of cutting edge techniques and concepts such as natural language processing and machine learning. | 0 | false | accepted_unique_top | 2019-03-01T20:59:07Z | donkey | https://ai.stackexchange.com/a/10952 | CC BY-SA 4.0 |
ai.stackexchange.com:q:10947 | [TITLE]
Does IBM Watson use machine learning?
[QUESTION]
I was reading an article on
Medium
and wanted to make it clear whether a bot created on IBM Watson is an intelligent one or unintelligent.
Simply put, there are 2 types of chatbots — unintelligent ones that act using predefined conversation flows (algorithms) wr... | ai.stackexchange.com:a:11044 | Yes, it does and at many parts of the solution. For one of the core components - intent detection - Intento did
a benchmark comparing IBM Watson and other similar products
.
Outside of intent detection, there are other areas where AI techniques help - e.g. disambiguation, bootstrapping a bot from chat logs etc. Specif... | 2 | true | accepted_unique_top | 2023-03-16T05:25:02Z | aameek | https://ai.stackexchange.com/a/11044 | CC BY-SA 4.0 |
ai.stackexchange.com:q:10950 | [TITLE]
Could error surface shape be useful to detect which local minima is better for generalization?
[QUESTION]
The following plot shows error function output based on system weights.
Two equal local minima are shown in green pointers. Note that the red dots are not related to the question.
Does the right one genera... | ai.stackexchange.com:a:10956 | I don't think that the concept of
generalization
is (directly) related to the "shape" of the function close to the point where it attains a minimum.
The concept of generalisation refers to when a trained model is able "
perform
well" on unseen data (that is, data not seen during the training phase). If a trained model... | 2 | false | accepted_unique_top | 2019-03-02T10:37:56Z | nbro | https://ai.stackexchange.com/a/10956 | CC BY-SA 4.0 |
ai.stackexchange.com:q:10950 | [TITLE]
Could error surface shape be useful to detect which local minima is better for generalization?
[QUESTION]
The following plot shows error function output based on system weights.
Two equal local minima are shown in green pointers. Note that the red dots are not related to the question.
Does the right one genera... | ai.stackexchange.com:a:10962 | In general I agree with @nbro answer, nevertheless sticking strictly to this specific question I'd like to share some speculations:
- what the author of the question provides us with is the Loss Function Shape so I'll try to use the full information here to compare the 2 minima
- looking at the LF steepness we observ... | 4 | true | accepted_unique_top | 2019-03-02T08:25:17Z | Nicola Bernini | https://ai.stackexchange.com/a/10962 | CC BY-SA 4.0 |
ai.stackexchange.com:q:10958 | [TITLE]
Can a sentence have different parse trees?
[QUESTION]
I just read about the concept of a
parse tree
.
In my understanding, a valid parse tree of a sentence needs to be validated by a linguistic expert. So, I concluded, a sentence only has one parse tree.
But, is that correct? Is it possible a sentence has mor... | ai.stackexchange.com:a:10960 | Grammars in NLP basically correspond to
Context-free Grammars(CFG)
in formal Language theory. And, in case the CFG corresponding to the NLP task is
ambiguous
, then corresponding to a single sentence (more formally
derivation
), there can be multiple Parse Trees.
Hence, it depends on the grammar whether there can be m... | 1 | false | accepted_unique_top | 2019-03-02T06:56:22Z | programmer | https://ai.stackexchange.com/a/10960 | CC BY-SA 4.0 |
ai.stackexchange.com:q:10958 | [TITLE]
Can a sentence have different parse trees?
[QUESTION]
I just read about the concept of a
parse tree
.
In my understanding, a valid parse tree of a sentence needs to be validated by a linguistic expert. So, I concluded, a sentence only has one parse tree.
But, is that correct? Is it possible a sentence has mor... | ai.stackexchange.com:a:10964 | But, is that correct? Is it possible a sentence has more than one valid parse tree (e.g. constituency-based)?
The fact that a single sequence of words can be parsed in different ways depending on context (or "grounding") is a common basis of miscommunication, misunderstanding, innuendo and jokes.
One classic NLP-rela... | 4 | true | accepted_unique_top | 2021-01-18T15:48:30Z | Neil Slater | https://ai.stackexchange.com/a/10964 | CC BY-SA 4.0 |
ai.stackexchange.com:q:10975 | [TITLE]
Why is MSE used over other quadratic loss functions?
[QUESTION]
So I was wondering, why I have only encountered square loss function also known as MSE. The only nice property of MSE I am so far aware of is its convex nature. But then all equations of the form
$x^{2n}$
where
$n$
is an integer belongs to the same... | ai.stackexchange.com:a:10976 | There is another variant for MSE. You can employ the absolute value of the difference of your hypothesis and the expected output.
`MSE`
and the absolute difference version, each have a property. The interpretation of the
`MSE`
is that you have the area of squares which at first are large but after training the predicte... | -1 | false | accepted_unique_top | 2019-03-03T09:35:53Z | Green Falcon | https://ai.stackexchange.com/a/10976 | CC BY-SA 4.0 |
ai.stackexchange.com:q:10975 | [TITLE]
Why is MSE used over other quadratic loss functions?
[QUESTION]
So I was wondering, why I have only encountered square loss function also known as MSE. The only nice property of MSE I am so far aware of is its convex nature. But then all equations of the form
$x^{2n}$
where
$n$
is an integer belongs to the same... | ai.stackexchange.com:a:15406 | I can comment on several properties of MSE and related losses.
As you mentioned MSE (aka
$l_2$
-loss) is convex which is a great property in optimization in which one can find a single global optimum. MSE is used in linear and non-linear least squares problems which form the basis of many widely used statistical metho... | 1 | true | accepted_unique_top | 2019-09-12T18:56:07Z | Anuar Y | https://ai.stackexchange.com/a/15406 | CC BY-SA 4.0 |
ai.stackexchange.com:q:11016 | [TITLE]
What does 'acting greedily' mean?
[QUESTION]
I wanted to clarify the term 'acting greedily'. What does it mean? Does it correspond to the immediate reward, future reward or both combined? I want to know the actions that will be taken in 2 cases:
- $v_\pi(s)$
is known and
$R_s$
is also known (only).
- $q_{\pi}... | ai.stackexchange.com:a:11017 | In general, a greedy "action" is an action that would lead to an immediate "benefit". For example, the
Dijkstra's algorithm
can be considered a greedy algorithm because at every step it selects the node with the smallest "estimate" to the initial (or starting) node. In reinforcement learning, a greedy action often refe... | 2 | false | accepted_unique_top | 2019-03-05T16:49:17Z | nbro | https://ai.stackexchange.com/a/11017 | CC BY-SA 4.0 |
ai.stackexchange.com:q:11016 | [TITLE]
What does 'acting greedily' mean?
[QUESTION]
I wanted to clarify the term 'acting greedily'. What does it mean? Does it correspond to the immediate reward, future reward or both combined? I want to know the actions that will be taken in 2 cases:
- $v_\pi(s)$
is known and
$R_s$
is also known (only).
- $q_{\pi}... | ai.stackexchange.com:a:11018 | In RL, the phrase "acting greedily" is usually short for "acting greedily with respect to the value function". Greedy local optimisation turns up in other contexts, and it is common to specify what metric is being maximised or minimised. The value function is most often the discounted sum of expected future reward, and... | 3 | true | accepted_unique_top | 2020-06-17T09:57:20Z | Neil Slater | https://ai.stackexchange.com/a/11018 | CC BY-SA 4.0 |
ai.stackexchange.com:q:11016 | [TITLE]
What does 'acting greedily' mean?
[QUESTION]
I wanted to clarify the term 'acting greedily'. What does it mean? Does it correspond to the immediate reward, future reward or both combined? I want to know the actions that will be taken in 2 cases:
- $v_\pi(s)$
is known and
$R_s$
is also known (only).
- $q_{\pi}... | ai.stackexchange.com:a:11023 | Acting greedily means that the search is not forward thinking and limits its decisions solely on immediate return. It is not quite the same as what is meant in human social contexts in that greed in that context can involve forward thinking strategies that sacrifice short term losses for long term gain. In the typical ... | 1 | false | accepted_unique_top | 2019-03-05T12:24:06Z | Douglas Daseeco | https://ai.stackexchange.com/a/11023 | CC BY-SA 4.0 |
ai.stackexchange.com:q:11043 | [TITLE]
Where can I find an implementation of the wake-sleep algorithm?
[QUESTION]
I'm looking to build from scratch an implementation of the
wake-sleep algorithm
, for unsupervised learning with neural networks. I plan on doing this in Python in order to better understand how it works. In order to facilitate my task, ... | ai.stackexchange.com:a:11051 | I don't know if you are looking for something in a library, but I've found
this
in a public Github (I've not checked deeply if it fits for you).
I hope that's what you're looking for. | 1 | false | accepted_unique_top | 2019-03-06T10:45:51Z | Angelo | https://ai.stackexchange.com/a/11051 | CC BY-SA 4.0 |
ai.stackexchange.com:q:11043 | [TITLE]
Where can I find an implementation of the wake-sleep algorithm?
[QUESTION]
I'm looking to build from scratch an implementation of the
wake-sleep algorithm
, for unsupervised learning with neural networks. I plan on doing this in Python in order to better understand how it works. In order to facilitate my task, ... | ai.stackexchange.com:a:12471 | I found the following detailed and well documented
Python notebook
, which uses only NumPy. | 3 | true | accepted_unique_top | 2019-11-18T18:54:56Z | TheCG | https://ai.stackexchange.com/a/12471 | CC BY-SA 4.0 |
ai.stackexchange.com:q:11074 | [TITLE]
Why does a fully connected layer only accept a fixed input size?
[QUESTION]
I'm studying how SPP (Spatial, Pyramid, Pooling) works. SPP was invented to tackle the fix input image size in CNN. According to the original paper
https://arxiv.org/pdf/1406.4729.pdf
, the authors say:
convolutional layers do not requ... | ai.stackexchange.com:a:11075 | A convolutional layer is a layer where you slide a kernel or filter (which you can think of as a small square matrix of weights, which need to be learned during the learning phase) over the input. In practice, when you need to slide this kernel, you will often need to specify the "padding" (around the input) and "strid... | 1 | true | accepted_unique_top | 2019-03-07T09:19:46Z | nbro | https://ai.stackexchange.com/a/11075 | CC BY-SA 4.0 |
ai.stackexchange.com:q:11074 | [TITLE]
Why does a fully connected layer only accept a fixed input size?
[QUESTION]
I'm studying how SPP (Spatial, Pyramid, Pooling) works. SPP was invented to tackle the fix input image size in CNN. According to the original paper
https://arxiv.org/pdf/1406.4729.pdf
, the authors say:
convolutional layers do not requ... | ai.stackexchange.com:a:11076 | It doesn't have to be so. Fully connected layer could be considered as convolutional layer with input image of 1 pixel and spatial kernel size of 1 pixel. So 1-pixel kernel convolutional layer is effectively the same as fully connected layer attached to each pixel. That is idea behind "Fully Convolutional Networks". If... | 0 | false | accepted_unique_top | 2019-03-07T10:28:09Z | mirror2image | https://ai.stackexchange.com/a/11076 | CC BY-SA 4.0 |
ai.stackexchange.com:q:11081 | [TITLE]
Difference in continuing and episodic cases in Sutton and Barto - Introduction to RL, exercise 3.5
[QUESTION]
Excercise 3.5 The equastions in Section 3.1 are for the continuing
case and need to be modified (very slightly) to apply to episodic
tasks. Show that you know the modifications needed by giving the
modi... | ai.stackexchange.com:a:11086 | Is it just about final states? So for
$s \in S$
when S is not final?
You are thinking the right way, but to represent what you mean you don't need to write out "when
$s$
is not final" - although that would be fine (and is used in some places), there is a more concise way of saying that given to you by the book.
As th... | 2 | true | accepted_unique_top | 2019-03-07T16:41:06Z | Neil Slater | https://ai.stackexchange.com/a/11086 | CC BY-SA 4.0 |
ai.stackexchange.com:q:11081 | [TITLE]
Difference in continuing and episodic cases in Sutton and Barto - Introduction to RL, exercise 3.5
[QUESTION]
Excercise 3.5 The equastions in Section 3.1 are for the continuing
case and need to be modified (very slightly) to apply to episodic
tasks. Show that you know the modifications needed by giving the
modi... | ai.stackexchange.com:a:11868 | I found myself turning in cycles for a while, so to clarify Neil Slater's answer,
In the beginning of the book,
$S$
means "set of non-terminal states" and
$S^+$
means "set of all states, including the terminal ones".
$$\sum_{s^{\prime} \in S} \sum_{r \in R} p(s^{\prime}, r | s,a) = 1, \forall s \in S, a \in A(s) \tag... | 0 | false | accepted_unique_top | 2019-04-17T18:23:17Z | Gigi | https://ai.stackexchange.com/a/11868 | CC BY-SA 4.0 |
ai.stackexchange.com:q:11166 | [TITLE]
What is geometric deep learning?
[QUESTION]
What is geometric deep learning (GDL)?
Here are a few sub-questions
- How is it different from deep learning?
- Why do we need GDL?
- What are some applications of GDL?
[TAGS]
deep-learning, definitions, geometric-deep-learning, graph-neural-networks, graphs | ai.stackexchange.com:a:11201 | The article
Geometric deep learning: going beyond Euclidean data
(by Michael M. Bronstein, Joan Bruna, Yann LeCun, Arthur Szlam, Pierre Vandergheynst) provides an overview of this relatively new sub-field of deep learning. It answers all the questions asked above (and more). If you are familiar with deep learning, grap... | 8 | true | accepted_below_top | 2021-02-03T01:27:19Z | nbro | https://ai.stackexchange.com/a/11201 | CC BY-SA 4.0 |
ai.stackexchange.com:q:11166 | [TITLE]
What is geometric deep learning?
[QUESTION]
What is geometric deep learning (GDL)?
Here are a few sub-questions
- How is it different from deep learning?
- Why do we need GDL?
- What are some applications of GDL?
[TAGS]
deep-learning, definitions, geometric-deep-learning, graph-neural-networks, graphs | ai.stackexchange.com:a:11252 | To complete the
first answer
that is rather graph oriented, I will write a little about deep learning on manifolds, which is quite general in terms of GDL thanks to the nature of manifolds.
Note that the description of GDL through the explanation of what are DL on graphs and manifolds, in opposition to DL on euclidean... | 20 | false | accepted_below_top | 2019-03-19T08:41:01Z | user22176 | https://ai.stackexchange.com/a/11252 | CC BY-SA 4.0 |
ai.stackexchange.com:q:11166 | [TITLE]
What is geometric deep learning?
[QUESTION]
What is geometric deep learning (GDL)?
Here are a few sub-questions
- How is it different from deep learning?
- Why do we need GDL?
- What are some applications of GDL?
[TAGS]
deep-learning, definitions, geometric-deep-learning, graph-neural-networks, graphs | ai.stackexchange.com:a:30369 | I didn't read the paper in depth, but one example of where assumptions of Euclidean space are made in the design of the networks are with ConvNets in image processing.
Specifically, Euclidean spaces are transformationally invariant, meaning that
$d(a,b) = d(a+c,b+c)$
. Each convolution layer iterates over the image wi... | 0 | false | accepted_below_top | 2021-08-25T15:45:22Z | k.c. sayz 'k.c sayz' | https://ai.stackexchange.com/a/30369 | CC BY-SA 4.0 |
ai.stackexchange.com:q:11182 | [TITLE]
Unable to understand the second iteration update in value iteration algorithm for solving MDP
[QUESTION]
I am trying to understand the value iteration method for Markov Decision Process(MDP) and I was referring to UC Berkeley's slides titled
Markov Decision Processes and Exact Solution Methods
On slide no. 9, ... | ai.stackexchange.com:a:11183 | First thing to know is that, in this case, values for the gridworld in new iteration are completely calculated with respect to the old values from the previous iteration. Value of
$0.78$
is got like this:
$0.9 \cdot (0.8 \cdot 1 + 0.1 \cdot 0.72 + 0.1 \cdot 0) = 0.7848 \approx 0.78$
term
$0.8 \cdot 1$
is for going to... | 2 | true | accepted_unique_top | 2019-03-12T22:05:07Z | Brale | https://ai.stackexchange.com/a/11183 | CC BY-SA 4.0 |
ai.stackexchange.com:q:11182 | [TITLE]
Unable to understand the second iteration update in value iteration algorithm for solving MDP
[QUESTION]
I am trying to understand the value iteration method for Markov Decision Process(MDP) and I was referring to UC Berkeley's slides titled
Markov Decision Processes and Exact Solution Methods
On slide no. 9, ... | ai.stackexchange.com:a:46142 | In second iteration,
using the bellman equation,
for (3,3), we got
0.8[0+(0.9)(1)]+0.1[0+(0.9)(0.72)]+0.1[0+(0.9)(0)] = 0.78
for (3,2), we got
0.8[0 + 0.9(0.72)] + 0.1[0 + 0.9(0)] + 0.1[0 + 0.9(-1)] = 0.43
for (2,3), we got
0.8[0 + 0.9(0.72)] + 0.1[0 + 0.9(0)] + 0.1[0 + 0.9(0)] = 0.52 | 0 | false | accepted_unique_top | 2024-07-03T07:07:11Z | Md. Foysal Ahmed | https://ai.stackexchange.com/a/46142 | CC BY-SA 4.0 |
ai.stackexchange.com:q:11219 | [TITLE]
Do we have to use CNN for Deep Q Learning?
[QUESTION]
I read top articles on Google Search about Deep Q-Learning:
- https://medium.freecodecamp.org/an-introduction-to-deep-q-learning-lets-play-doom-54d02d8017d8
- https://skymind.ai/wiki/deep-reinforcement-learning
- https://neuro.cs.ut.ee/demystifying-deep-r... | ai.stackexchange.com:a:11221 | No. DQN and other deep RL methods work well with fully connected layers. Here's an implementation of DQN which doesn't use CNNs:
github.com/keon/deep-q-learning/blob/master/dqn.py
DeepMind mostly use CNN because they use image as input state, and that because they tried to evaluate performance of their methods vs huma... | 10 | true | accepted_unique_top | 2019-03-15T13:38:47Z | mirror2image | https://ai.stackexchange.com/a/11221 | CC BY-SA 4.0 |
ai.stackexchange.com:q:11219 | [TITLE]
Do we have to use CNN for Deep Q Learning?
[QUESTION]
I read top articles on Google Search about Deep Q-Learning:
- https://medium.freecodecamp.org/an-introduction-to-deep-q-learning-lets-play-doom-54d02d8017d8
- https://skymind.ai/wiki/deep-reinforcement-learning
- https://neuro.cs.ut.ee/demystifying-deep-r... | ai.stackexchange.com:a:11308 | The approximator can be any artificial neural network architecture, including deep fully-connected networks. | 0 | false | accepted_unique_top | 2019-03-22T09:10:32Z | Bionic Buffulo | https://ai.stackexchange.com/a/11308 | CC BY-SA 4.0 |
ai.stackexchange.com:q:11226 | [TITLE]
What is non-Euclidean data?
[QUESTION]
What is non-Euclidean data?
Here are some sub-questions
- Where does this type of data arise? I have come across this term in the context of geometric deep learning and graph neural networks.
- Apparently, graphs and manifolds are non-Euclidean data. Why exactly is that... | ai.stackexchange.com:a:11259 | Non-Euclidian geometry can be generally boiled down to the phrase
the shortest path between 2 points isn't necessarily a straight line.
Or, put in a way that lends itself very much to machine learning,
things that are similar to each other are not necessarily close if one uses Euclidean distance as a metric (aka the... | 10 | false | accepted_unique_top | 2020-07-10T11:50:16Z | Jaden Travnik | https://ai.stackexchange.com/a/11259 | CC BY-SA 4.0 |
ai.stackexchange.com:q:11226 | [TITLE]
What is non-Euclidean data?
[QUESTION]
What is non-Euclidean data?
Here are some sub-questions
- Where does this type of data arise? I have come across this term in the context of geometric deep learning and graph neural networks.
- Apparently, graphs and manifolds are non-Euclidean data. Why exactly is that... | ai.stackexchange.com:a:20628 | I presume this question was prompted by the paper
Geometric deep learning:
going beyond Euclidean data
(2017). If we look at its abstract:
Many scientific fields study
data with an underlying structure that is a non-Euclidean space.
Some examples include social networks in computational social sciences, sensor network... | 16 | true | accepted_unique_top | 2020-04-25T18:59:57Z | End genocide - save Gaza | https://ai.stackexchange.com/a/20628 | CC BY-SA 4.0 |
ai.stackexchange.com:q:11226 | [TITLE]
What is non-Euclidean data?
[QUESTION]
What is non-Euclidean data?
Here are some sub-questions
- Where does this type of data arise? I have come across this term in the context of geometric deep learning and graph neural networks.
- Apparently, graphs and manifolds are non-Euclidean data. Why exactly is that... | ai.stackexchange.com:a:22432 | As far as I understand, the concept of
non-Euclidean space
doesn't bring the ordinality or hierarchy among the features, compared to that with the data formed in the Euclidean space.
The difference between both these techniques is not remarkable for discriminative tasks like classification. But, for generative modelin... | 0 | false | accepted_unique_top | 2020-07-10T11:51:44Z | Devansh Khandekar | https://ai.stackexchange.com/a/22432 | CC BY-SA 4.0 |
ai.stackexchange.com:q:11226 | [TITLE]
What is non-Euclidean data?
[QUESTION]
What is non-Euclidean data?
Here are some sub-questions
- Where does this type of data arise? I have come across this term in the context of geometric deep learning and graph neural networks.
- Apparently, graphs and manifolds are non-Euclidean data. Why exactly is that... | ai.stackexchange.com:a:30368 | It's hard to say because Euclidean space is defined with respect to some kind of metric, so without any clearer exposition on the nature of the data/problem, the phrase itself may or may not be clear.
A
metric
$d: A \times A \rightarrow \mathbb{R}$
is a function that defines distance between any two points in the spa... | 1 | false | accepted_unique_top | 2021-08-25T15:33:09Z | k.c. sayz 'k.c sayz' | https://ai.stackexchange.com/a/30368 | CC BY-SA 4.0 |
ai.stackexchange.com:q:11226 | [TITLE]
What is non-Euclidean data?
[QUESTION]
What is non-Euclidean data?
Here are some sub-questions
- Where does this type of data arise? I have come across this term in the context of geometric deep learning and graph neural networks.
- Apparently, graphs and manifolds are non-Euclidean data. Why exactly is that... | ai.stackexchange.com:a:35228 | Where does this type of data arise?
In terms of
learning
on non-Euclidean data, it was probably first coined by prof. Bronstein
here
. Recently, prof. Bronstein published, along with other top authors in the field, a
book
about geometric deep learning, which in essence presents a unified mathematical framework for sym... | 1 | false | accepted_unique_top | 2022-04-16T20:43:02Z | Andrei-Cristian Rad | https://ai.stackexchange.com/a/35228 | CC BY-SA 4.0 |
ai.stackexchange.com:q:11285 | [TITLE]
What is the difference between latent and embedding spaces?
[QUESTION]
In general, the word "latent" means "hidden" and "to embed" means "to incorporate". In machine learning, the expressions "hidden (or latent) space" and "embedding space" occur in several contexts. More specifically, an
embedding
can refer to... | ai.stackexchange.com:a:11287 | When it comes to normal layman terms "latent space" means it cannot be accessed, thus we have no direct control over it. We can only manipulate it indirectly, while "Embeddings" can be obtained directly. We can use deterministic operations or transformations to convert an object into its corresponding embedding space.
... | 11 | false | accepted_unique_top | 2019-05-24T12:19:18Z | user9947 | https://ai.stackexchange.com/a/11287 | CC BY-SA 4.0 |
ai.stackexchange.com:q:11285 | [TITLE]
What is the difference between latent and embedding spaces?
[QUESTION]
In general, the word "latent" means "hidden" and "to embed" means "to incorporate". In machine learning, the expressions "hidden (or latent) space" and "embedding space" occur in several contexts. More specifically, an
embedding
can refer to... | ai.stackexchange.com:a:12507 | The expression "latent space" explicitly indicates that the space is associated with the mathematical concept of an hidden (or latent) variable, which cannot be observed directly, but only indirectly.
The expression "embedding space" refers to a vector space that represents an original space of inputs (e.g. images or ... | 6 | false | accepted_unique_top | 2019-05-24T12:22:09Z | nbro | https://ai.stackexchange.com/a/12507 | CC BY-SA 4.0 |
ai.stackexchange.com:q:11285 | [TITLE]
What is the difference between latent and embedding spaces?
[QUESTION]
In general, the word "latent" means "hidden" and "to embed" means "to incorporate". In machine learning, the expressions "hidden (or latent) space" and "embedding space" occur in several contexts. More specifically, an
embedding
can refer to... | ai.stackexchange.com:a:20646 | Embedding vs Latent Space
Due to Machine Learning's recent and rapid renaissance, and the fact that it draws from many distinct areas of mathematics, statistics, and computer science, it often has a number of different terms for the same or similar concepts.
"Latent space" and "embedding" both refer to an (often lowe... | 38 | true | accepted_unique_top | 2020-12-31T17:43:22Z | End genocide - save Gaza | https://ai.stackexchange.com/a/20646 | CC BY-SA 4.0 |
ai.stackexchange.com:q:11285 | [TITLE]
What is the difference between latent and embedding spaces?
[QUESTION]
In general, the word "latent" means "hidden" and "to embed" means "to incorporate". In machine learning, the expressions "hidden (or latent) space" and "embedding space" occur in several contexts. More specifically, an
embedding
can refer to... | ai.stackexchange.com:a:20712 | The words latent space and embedding space are often used interchangeably. However, latent space can more specifically refer to the sample space of a stochastic representation, whereas embedding space more often refers to the space of a deterministic representation.
This comes from latent referring to an unobserved ra... | 0 | false | accepted_unique_top | 2020-04-28T04:00:17Z | danijar | https://ai.stackexchange.com/a/20712 | CC BY-SA 4.0 |
ai.stackexchange.com:q:11285 | [TITLE]
What is the difference between latent and embedding spaces?
[QUESTION]
In general, the word "latent" means "hidden" and "to embed" means "to incorporate". In machine learning, the expressions "hidden (or latent) space" and "embedding space" occur in several contexts. More specifically, an
embedding
can refer to... | ai.stackexchange.com:a:28319 | To give a statistician's answer, the distinction is empirical (embedding) versus theoretical (latent positions). You define a statistical model which has latent positions that you could then try to estimate, given data. Or, given data, you might simply find a vector representation of each object of interest in a way th... | 1 | false | accepted_unique_top | 2021-06-18T14:25:33Z | PRD | https://ai.stackexchange.com/a/28319 | CC BY-SA 4.0 |
ai.stackexchange.com:q:11293 | [TITLE]
What determines the values of weights in a neural network?
[QUESTION]
I am trying to understand how weights are actually gotten. What is generating them in a neural network? What is the algorithm that gives them certain values?
[TAGS]
machine-learning, neural-networks, neurons | ai.stackexchange.com:a:11294 | Typically, weights are randomly initialized. Then, as the model is
optimized
for its given task, those weights are steadily made "better" as determined by the network's
loss function
. This is also referred to as "training" the neural network.
By far the most popular way of updating weights in a neural net is the
back... | 5 | true | accepted_unique_top | 2019-03-17T19:55:17Z | Philip Raeisghasem | https://ai.stackexchange.com/a/11294 | CC BY-SA 4.0 |
ai.stackexchange.com:q:11293 | [TITLE]
What determines the values of weights in a neural network?
[QUESTION]
I am trying to understand how weights are actually gotten. What is generating them in a neural network? What is the algorithm that gives them certain values?
[TAGS]
machine-learning, neural-networks, neurons | ai.stackexchange.com:a:11295 | I agree with @PhilipRaeisghasem, in most architectures, weights are initialized in a random manner. However, some research papers suggest applying a random normal distribution initialization to the weights in the case of Convolutional Neural Networks (for computer vision). | 0 | false | accepted_unique_top | 2019-03-17T21:44:44Z | iustin | https://ai.stackexchange.com/a/11295 | CC BY-SA 4.0 |
ai.stackexchange.com:q:113 | [TITLE]
What's the difference between hyperbolic tangent and sigmoid neurons?
[QUESTION]
Two common activation functions used in deep learning are the hyperbolic tangent function and the sigmoid activation function. I understand that the hyperbolic tangent is just a rescaling and translation of the sigmoid function:
$... | ai.stackexchange.com:a:119 | I don't think it makes sense to decide activation functions based on desired properties of the output; you can easily insert a calibration step that maps the 'neural network score' to whatever units you actually want to use (dollars, probability, etc.).
So I think preference between different activation functions most... | 3 | false | accepted_unique_top | 2016-08-02T19:28:00Z | Matthew Gray | https://ai.stackexchange.com/a/119 | CC BY-SA 3.0 |
ai.stackexchange.com:q:113 | [TITLE]
What's the difference between hyperbolic tangent and sigmoid neurons?
[QUESTION]
Two common activation functions used in deep learning are the hyperbolic tangent function and the sigmoid activation function. I understand that the hyperbolic tangent is just a rescaling and translation of the sigmoid function:
$... | ai.stackexchange.com:a:5572 | Sigmoid > Hyperbolic tangent:
As you mentioned, the application of Sigmoid might be more convenient than hyperbolic tangent in the cases that we need a probability value at the output (as @matthew-graves says, we can fix this with a simple mapping/calibration step). In other layers, this makes no sense.
Hyperbolic ta... | 5 | true | accepted_unique_top | 2019-04-02T18:31:34Z | Borhan Kazimipour | https://ai.stackexchange.com/a/5572 | CC BY-SA 4.0 |
ai.stackexchange.com:q:11342 | [TITLE]
How to chose dense layer size?
[QUESTION]
I am fine-tuning a VGG16 model on 20 classes with 500k images I was wondering how do you chose the size of the dense layer (the one before the prediction layer which has a size 20). I would prefer not to do a grid search seeing how long it take to train my model.
Also ... | ai.stackexchange.com:a:11343 | It's depend more on number of classes. For 20 classes 2 layers 512 should be more then enough. If you want to experiment you can try also 2 x 256 and 2 x 1024. Less then 256 may work too, but you may underutilize power of previous conv layers. | 3 | true | accepted_tied_top | 2019-03-20T11:20:55Z | mirror2image | https://ai.stackexchange.com/a/11343 | CC BY-SA 4.0 |
ai.stackexchange.com:q:11342 | [TITLE]
How to chose dense layer size?
[QUESTION]
I am fine-tuning a VGG16 model on 20 classes with 500k images I was wondering how do you chose the size of the dense layer (the one before the prediction layer which has a size 20). I would prefer not to do a grid search seeing how long it take to train my model.
Also ... | ai.stackexchange.com:a:26438 | I am also wondering about this.
It must depend both on convolutional sub-network output size (N) and number of classes (M).
Maybe there are some rules of thumbs depending on (N, M).
- Why 2 dense layers and not, say, 3 or 4 ?
- Is it better to have all dense layers (except last) the same size ? or decreasing ? or inc... | 3 | false | accepted_tied_top | 2021-02-18T12:17:39Z | lostdatum | https://ai.stackexchange.com/a/26438 | CC BY-SA 4.0 |
ai.stackexchange.com:q:11350 | [TITLE]
Reinforcement Learning with long term rewards and fixed states and actions
[QUESTION]
I have read a lot about RL algorithms, that update the action-value function at each step with the currently gained reward. The requirement here is, that the reward is obtained after each step.
I have a case, where I have thr... | ai.stackexchange.com:a:11351 | You don't need to have a reward on every single timestep, reward at the end is enough. Reinforcement learning can deal with temporal credit assignment problem, all algorithms are designed to work with it. Its enough to define a reward at the end where you, for example, give a reward of 1 if sentence is satisfactory or ... | 4 | false | accepted_tied_top | 2019-03-20T22:49:12Z | Brale | https://ai.stackexchange.com/a/11351 | CC BY-SA 4.0 |
ai.stackexchange.com:q:11350 | [TITLE]
Reinforcement Learning with long term rewards and fixed states and actions
[QUESTION]
I have read a lot about RL algorithms, that update the action-value function at each step with the currently gained reward. The requirement here is, that the reward is obtained after each step.
I have a case, where I have thr... | ai.stackexchange.com:a:11352 | You are describing a straightforward Markov Decision Process that could be solved by almost any Reinforcement Learning algorithm.
I have read a lot about RL algorithms, that update the action-value function at each step with the currently gained reward. The requirement here is, that the reward is obtained after each s... | 4 | true | accepted_tied_top | 2019-03-21T08:11:38Z | Neil Slater | https://ai.stackexchange.com/a/11352 | CC BY-SA 4.0 |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.