tuning a convolution neural net, sample sizeSignal classification with convolution neural networkConvnet training error does not decreaseRelation between convolution in math and CNNHow to use the output of GridSearch?How to input & pre-process images for a Deep Convolutional Neural Network?Automated tuning of HyperparameterConvolution over volume in CNNsunderstanding the filter function in convolution neural networkTest Loss plateau fast in Convolutional Neural NetDisadvantages of hyperparameter tuning on a random sample of dataset
Does a large simulator bay have standard public address announcements?
How can I practically buy stocks?
Is there any official lore on the Far Realm?
How does Captain America channel this power?
How exactly does Hawking radiation decrease the mass of black holes?
How can I print the prosodic symbols in LaTeX?
What makes accurate emulation of old systems a difficult task?
What's the polite way to say "I need to urinate"?
What is the optimal strategy for the Dictionary Game?
Can someone publish a story that happened to you?
'It addicted me, with one taste.' Can 'addict' be used transitively?
Why did C use the -> operator instead of reusing the . operator?
How would 10 generations of living underground change the human body?
If a planet has 3 moons, is it possible to have triple Full/New Moons at once?
Why was the Spitfire's elliptical wing almost uncopied by other aircraft of World War 2?
Random Forest different results for same observation
How do I deal with a coworker that keeps asking to make small superficial changes to a report, and it is seriously triggering my anxiety?
How much cash can I safely carry into the USA and avoid civil forfeiture?
Your bread will be buttered on both sides
Why must Chinese maps be obfuscated?
Apparently, my CLR assembly function is causing deadlocks?
How to stop co-workers from teasing me because I know Russian?
Do I have an "anti-research" personality?
Re-entry to Germany after vacation using blue card
tuning a convolution neural net, sample size
Signal classification with convolution neural networkConvnet training error does not decreaseRelation between convolution in math and CNNHow to use the output of GridSearch?How to input & pre-process images for a Deep Convolutional Neural Network?Automated tuning of HyperparameterConvolution over volume in CNNsunderstanding the filter function in convolution neural networkTest Loss plateau fast in Convolutional Neural NetDisadvantages of hyperparameter tuning on a random sample of dataset
$begingroup$
I keep reading that convolution neural net (CNN) performs best with lots and lots (100k+) of data. Is there any rule of thumb, or lower limit for data size during the grid search phase?
For example, if I run a CNN with 100 data points, vary just one parameter (say add an extra layer, or increase a filter size), and get better results, can I reasonably expect better results with those parameters during the actual training phase?
machine-learning cnn convolution hyperparameter-tuning
asked Apr 4 '18 at 15:41
Mohammad AtharMohammad Athar
261111
$endgroup$
add a comment |
$begingroup$
I keep reading that convolution neural net (CNN) performs best with lots and lots (100k+) of data. Is there any rule of thumb, or lower limit for data size during the grid search phase?
For example, if I run a CNN with 100 data points, vary just one parameter (say add an extra layer, or increase a filter size), and get better results, can I reasonably expect better results with those parameters during the actual training phase?
machine-learning cnn convolution hyperparameter-tuning
asked Apr 4 '18 at 15:41
Mohammad AtharMohammad Athar
261111
$endgroup$
$begingroup$
It's wrong that you need a lot of images to train a conv-net... I had only trained them with 22 images as trainset and 7 as validation and test.... that also works
$endgroup$
– Aditya
Apr 5 '18 at 2:35
add a comment |
$begingroup$
I keep reading that convolution neural net (CNN) performs best with lots and lots (100k+) of data. Is there any rule of thumb, or lower limit for data size during the grid search phase?
For example, if I run a CNN with 100 data points, vary just one parameter (say add an extra layer, or increase a filter size), and get better results, can I reasonably expect better results with those parameters during the actual training phase?
machine-learning cnn convolution hyperparameter-tuning
asked Apr 4 '18 at 15:41
Mohammad AtharMohammad Athar
261111
$endgroup$
I keep reading that convolution neural net (CNN) performs best with lots and lots (100k+) of data. Is there any rule of thumb, or lower limit for data size during the grid search phase?
For example, if I run a CNN with 100 data points, vary just one parameter (say add an extra layer, or increase a filter size), and get better results, can I reasonably expect better results with those parameters during the actual training phase?
machine-learning cnn convolution hyperparameter-tuning
machine-learning cnn convolution hyperparameter-tuning
asked Apr 4 '18 at 15:41
Mohammad AtharMohammad Athar
261111
asked Apr 4 '18 at 15:41
Mohammad AtharMohammad Athar
261111
asked Apr 4 '18 at 15:41
Mohammad AtharMohammad Athar
261111
asked Apr 4 '18 at 15:41
Mohammad AtharMohammad Athar
261111
asked Apr 4 '18 at 15:41
Mohammad AtharMohammad Athar
261111
261111
$begingroup$
It's wrong that you need a lot of images to train a conv-net... I had only trained them with 22 images as trainset and 7 as validation and test.... that also works
$endgroup$
– Aditya
Apr 5 '18 at 2:35
add a comment |
$begingroup$
It's wrong that you need a lot of images to train a conv-net... I had only trained them with 22 images as trainset and 7 as validation and test.... that also works
$endgroup$
– Aditya
Apr 5 '18 at 2:35
$begingroup$
It's wrong that you need a lot of images to train a conv-net... I had only trained them with 22 images as trainset and 7 as validation and test.... that also works
$endgroup$
– Aditya
Apr 5 '18 at 2:35
$begingroup$
It's wrong that you need a lot of images to train a conv-net... I had only trained them with 22 images as trainset and 7 as validation and test.... that also works
$endgroup$
– Aditya
Apr 5 '18 at 2:35
add a comment |
1 Answer
1
active
oldest
votes
$begingroup$
If you use pre-trained weights, you need significantly lesser data as the initial layers have already learned from a ton of data and you just need to fine tune the later ones.
What you said is not true, you can train on CIFAR10 and get 90%+ and that is not 100k+. It depends on the complexity of the data and how similar the features are. If they are easily Seperable -less data. If the disctintions are harder then the model needs a lot of examples to figure out which of features are seperate.
I would say you could IF you sample was representive of the population.
answered Nov 7 '18 at 17:04
Rahul DeoraRahul Deora
1
$endgroup$
add a comment |
Your Answer
StackExchange.ready(function()
var channelOptions =
tags: "".split(" "),
id: "557"
;
initTagRenderer("".split(" "), "".split(" "), channelOptions);
StackExchange.using("externalEditor", function()
// Have to fire editor after snippets, if snippets enabled
if (StackExchange.settings.snippets.snippetsEnabled)
StackExchange.using("snippets", function()
createEditor();
);
else
createEditor();
);
function createEditor()
StackExchange.prepareEditor(
heartbeatType: 'answer',
autoActivateHeartbeat: false,
convertImagesToLinks: false,
noModals: true,
showLowRepImageUploadWarning: true,
reputationToPostImages: null,
bindNavPrevention: true,
postfix: "",
imageUploader:
brandingHtml: "Powered by u003ca class="icon-imgur-white" href="https://imgur.com/"u003eu003c/au003e",
contentPolicyHtml: "User contributions licensed under u003ca href="https://creativecommons.org/licenses/by-sa/3.0/"u003ecc by-sa 3.0 with attribution requiredu003c/au003e u003ca href="https://stackoverflow.com/legal/content-policy"u003e(content policy)u003c/au003e",
allowUrls: true
,
onDemand: true,
discardSelector: ".discard-answer"
,immediatelyShowMarkdownHelp:true
);
);
Sign up or log in
StackExchange.ready(function ()
StackExchange.helpers.onClickDraftSave('#login-link');
);
Sign up using Google
Sign up using Facebook
Sign up using Email and Password
Post as a guest
Required, but never shown
StackExchange.ready(
function ()
StackExchange.openid.initPostLogin('.new-post-login', 'https%3a%2f%2fdatascience.stackexchange.com%2fquestions%2f29904%2ftuning-a-convolution-neural-net-sample-size%23new-answer', 'question_page');
);
Post as a guest
Required, but never shown
1 Answer
1
active
oldest
votes
1 Answer
1
active
oldest
votes
active
oldest
votes
active
oldest
votes
$begingroup$
If you use pre-trained weights, you need significantly lesser data as the initial layers have already learned from a ton of data and you just need to fine tune the later ones.
What you said is not true, you can train on CIFAR10 and get 90%+ and that is not 100k+. It depends on the complexity of the data and how similar the features are. If they are easily Seperable -less data. If the disctintions are harder then the model needs a lot of examples to figure out which of features are seperate.
I would say you could IF you sample was representive of the population.
answered Nov 7 '18 at 17:04
Rahul DeoraRahul Deora
1
$endgroup$
add a comment |
$begingroup$
If you use pre-trained weights, you need significantly lesser data as the initial layers have already learned from a ton of data and you just need to fine tune the later ones.
What you said is not true, you can train on CIFAR10 and get 90%+ and that is not 100k+. It depends on the complexity of the data and how similar the features are. If they are easily Seperable -less data. If the disctintions are harder then the model needs a lot of examples to figure out which of features are seperate.
I would say you could IF you sample was representive of the population.
answered Nov 7 '18 at 17:04
Rahul DeoraRahul Deora
1
$endgroup$
add a comment |
$begingroup$
If you use pre-trained weights, you need significantly lesser data as the initial layers have already learned from a ton of data and you just need to fine tune the later ones.
What you said is not true, you can train on CIFAR10 and get 90%+ and that is not 100k+. It depends on the complexity of the data and how similar the features are. If they are easily Seperable -less data. If the disctintions are harder then the model needs a lot of examples to figure out which of features are seperate.
I would say you could IF you sample was representive of the population.
answered Nov 7 '18 at 17:04
Rahul DeoraRahul Deora
1
$endgroup$
If you use pre-trained weights, you need significantly lesser data as the initial layers have already learned from a ton of data and you just need to fine tune the later ones.
What you said is not true, you can train on CIFAR10 and get 90%+ and that is not 100k+. It depends on the complexity of the data and how similar the features are. If they are easily Seperable -less data. If the disctintions are harder then the model needs a lot of examples to figure out which of features are seperate.
I would say you could IF you sample was representive of the population.
answered Nov 7 '18 at 17:04
Rahul DeoraRahul Deora
1
answered Nov 7 '18 at 17:04
Rahul DeoraRahul Deora
1
answered Nov 7 '18 at 17:04
Rahul DeoraRahul Deora
1
answered Nov 7 '18 at 17:04
Rahul DeoraRahul Deora
1
1
add a comment |
add a comment |
Thanks for contributing an answer to Data Science Stack Exchange!
- Please be sure to answer the question. Provide details and share your research!
But avoid …
- Asking for help, clarification, or responding to other answers.
- Making statements based on opinion; back them up with references or personal experience.
Use MathJax to format equations. MathJax reference.
To learn more, see our tips on writing great answers.
Sign up or log in
StackExchange.ready(function ()
StackExchange.helpers.onClickDraftSave('#login-link');
);
Sign up using Google
Sign up using Facebook
Sign up using Email and Password
Post as a guest
Required, but never shown
StackExchange.ready(
function ()
StackExchange.openid.initPostLogin('.new-post-login', 'https%3a%2f%2fdatascience.stackexchange.com%2fquestions%2f29904%2ftuning-a-convolution-neural-net-sample-size%23new-answer', 'question_page');
);
Post as a guest
Required, but never shown
Sign up or log in
StackExchange.ready(function ()
StackExchange.helpers.onClickDraftSave('#login-link');
);
Sign up using Google
Sign up using Facebook
Sign up using Email and Password
Post as a guest
Required, but never shown
Sign up or log in
StackExchange.ready(function ()
StackExchange.helpers.onClickDraftSave('#login-link');
);
Sign up using Google
Sign up using Facebook
Sign up using Email and Password
Post as a guest
Required, but never shown
Sign up or log in
StackExchange.ready(function ()
StackExchange.helpers.onClickDraftSave('#login-link');
);
Sign up using Google
Sign up using Facebook
Sign up using Email and Password
Sign up using Google
Sign up using Facebook
Sign up using Email and Password
Post as a guest
Required, but never shown
Required, but never shown
Required, but never shown
Required, but never shown
Required, but never shown
Required, but never shown
Required, but never shown
Required, but never shown
Required, but never shown
$begingroup$
It's wrong that you need a lot of images to train a conv-net... I had only trained them with 22 images as trainset and 7 as validation and test.... that also works
$endgroup$
– Aditya
Apr 5 '18 at 2:35