Downsampling the dataset to create balanced dataset for neural modelsShould I go for a 'balanced' dataset or a 'representative' dataset?Clustering of numerical dataConvolution Neural Network Loss and performanceCreating Balanced Dataset Using ScikitsTraining Error decreasing with each epochHow to create pipeline for Neural Network models?Create a binary-classification dataset (python: sklearn.datasets.make_classification)How to interpret Sum of squared errorTrain classifier on balanced dataset and apply on imbalanced dataset?Using SMOTE for Synthetic Data generation to improve performance on unbalanced data
What happens to Mjolnir (Thor's hammer) at the end of Endgame?
Dynamic SOQL query relationship with field visibility for Users
Why did some of my point & shoot film photos come back with one third light white or orange?
On The Origin of Dissonant Chords
What term is being referred to with "reflected-sound-of-underground-spirits"?
Checks user level and limit the data before saving it to mongoDB
"You've called the wrong number" or "You called the wrong number"
can anyone help me with this awful query plan?
"Hidden" theta-term in Hamiltonian formulation of Yang-Mills theory
Why does Mind Blank stop the Feeblemind spell?
Two field separators (colon and space) in awk
Does a large simulator bay have standard public address announcements?
What happened to Captain America in Endgame?
"Whatever a Russian does, they end up making the Kalashnikov gun"? Are there any similar proverbs in English?
Minor Revision with suggestion of an alternative proof by reviewer
Rivers without rain
How to denote matrix elements succinctly?
Coordinate my way to the name of the (video) game
Don’t seats that recline flat defeat the purpose of having seatbelts?
Is there a way to generate a list of distinct numbers such that no two subsets ever have an equal sum?
What does ゆーか mean?
Can I criticise the more senior developers around me for not writing clean code?
What is the philosophical significance of speech acts/implicature?
Phrase for the opposite of "foolproof"
Downsampling the dataset to create balanced dataset for neural models
Should I go for a 'balanced' dataset or a 'representative' dataset?Clustering of numerical dataConvolution Neural Network Loss and performanceCreating Balanced Dataset Using ScikitsTraining Error decreasing with each epochHow to create pipeline for Neural Network models?Create a binary-classification dataset (python: sklearn.datasets.make_classification)How to interpret Sum of squared errorTrain classifier on balanced dataset and apply on imbalanced dataset?Using SMOTE for Synthetic Data generation to improve performance on unbalanced data
$begingroup$
I have a classification dataset with 10k instances and 4 classes and it is unbalanced. 7000 of it belongs to first class, 2000 of it belongs to second 800 of it belongs to third class and remaining instances are belong to fourth class. I have a single neural network based on MLP.
At the end of down-sampling process, I am planning to have 2000,2000,800,200 instances. (I can even further down-sample first 3 classes until I have 200 instances from each class but I don't want to decrease size of the dataset that much).
My question is: If I train my MLP for N epoch, should I use the same subset for class-1 for all of the epochs or should I generate this subset randomly for each epoch?
neural-network scikit-learn dataset sampling
asked Sep 26 '18 at 9:34
zwlayerzwlayer
1061
$endgroup$
bumped to the homepage by Community♦ 10 mins ago
This question has answers that may be good or bad; the system has marked it active so that they can be reviewed.
add a comment |
$begingroup$
I have a classification dataset with 10k instances and 4 classes and it is unbalanced. 7000 of it belongs to first class, 2000 of it belongs to second 800 of it belongs to third class and remaining instances are belong to fourth class. I have a single neural network based on MLP.
At the end of down-sampling process, I am planning to have 2000,2000,800,200 instances. (I can even further down-sample first 3 classes until I have 200 instances from each class but I don't want to decrease size of the dataset that much).
My question is: If I train my MLP for N epoch, should I use the same subset for class-1 for all of the epochs or should I generate this subset randomly for each epoch?
neural-network scikit-learn dataset sampling
asked Sep 26 '18 at 9:34
zwlayerzwlayer
1061
$endgroup$
bumped to the homepage by Community♦ 10 mins ago
This question has answers that may be good or bad; the system has marked it active so that they can be reviewed.
add a comment |
$begingroup$
I have a classification dataset with 10k instances and 4 classes and it is unbalanced. 7000 of it belongs to first class, 2000 of it belongs to second 800 of it belongs to third class and remaining instances are belong to fourth class. I have a single neural network based on MLP.
At the end of down-sampling process, I am planning to have 2000,2000,800,200 instances. (I can even further down-sample first 3 classes until I have 200 instances from each class but I don't want to decrease size of the dataset that much).
My question is: If I train my MLP for N epoch, should I use the same subset for class-1 for all of the epochs or should I generate this subset randomly for each epoch?
neural-network scikit-learn dataset sampling
asked Sep 26 '18 at 9:34
zwlayerzwlayer
1061
$endgroup$
I have a classification dataset with 10k instances and 4 classes and it is unbalanced. 7000 of it belongs to first class, 2000 of it belongs to second 800 of it belongs to third class and remaining instances are belong to fourth class. I have a single neural network based on MLP.
At the end of down-sampling process, I am planning to have 2000,2000,800,200 instances. (I can even further down-sample first 3 classes until I have 200 instances from each class but I don't want to decrease size of the dataset that much).
My question is: If I train my MLP for N epoch, should I use the same subset for class-1 for all of the epochs or should I generate this subset randomly for each epoch?
neural-network scikit-learn dataset sampling
neural-network scikit-learn dataset sampling
asked Sep 26 '18 at 9:34
zwlayerzwlayer
1061
asked Sep 26 '18 at 9:34
zwlayerzwlayer
1061
asked Sep 26 '18 at 9:34
zwlayerzwlayer
1061
asked Sep 26 '18 at 9:34
zwlayerzwlayer
1061
asked Sep 26 '18 at 9:34
zwlayerzwlayer
1061
1061
bumped to the homepage by Community♦ 10 mins ago
This question has answers that may be good or bad; the system has marked it active so that they can be reviewed.
bumped to the homepage by Community♦ 10 mins ago
This question has answers that may be good or bad; the system has marked it active so that they can be reviewed.
add a comment |
add a comment |
1 Answer
1
active
oldest
votes
$begingroup$
You will get better results if you change subset of images from downsampled class randomly at each epoch. It will help in building a robust model as well. And use data augmentation techniques for minor classes.
answered Sep 26 '18 at 10:24
Poonam LigadePoonam Ligade
413
$endgroup$
$begingroup$
I guess you missed my question. I am not asking how to balanced my dataset, I am asking should I changed my down-sampled class at each epoch or keep it same during traning of my neural net.
$endgroup$
– zwlayer
Sep 26 '18 at 11:02
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%2f38805%2fdownsampling-the-dataset-to-create-balanced-dataset-for-neural-models%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
$begingroup$
You will get better results if you change subset of images from downsampled class randomly at each epoch. It will help in building a robust model as well. And use data augmentation techniques for minor classes.
answered Sep 26 '18 at 10:24
Poonam LigadePoonam Ligade
413
$endgroup$
$begingroup$
I guess you missed my question. I am not asking how to balanced my dataset, I am asking should I changed my down-sampled class at each epoch or keep it same during traning of my neural net.
$endgroup$
– zwlayer
Sep 26 '18 at 11:02
add a comment |
$begingroup$
You will get better results if you change subset of images from downsampled class randomly at each epoch. It will help in building a robust model as well. And use data augmentation techniques for minor classes.
answered Sep 26 '18 at 10:24
Poonam LigadePoonam Ligade
413
$endgroup$
$begingroup$
I guess you missed my question. I am not asking how to balanced my dataset, I am asking should I changed my down-sampled class at each epoch or keep it same during traning of my neural net.
$endgroup$
– zwlayer
Sep 26 '18 at 11:02
add a comment |
$begingroup$
You will get better results if you change subset of images from downsampled class randomly at each epoch. It will help in building a robust model as well. And use data augmentation techniques for minor classes.
answered Sep 26 '18 at 10:24
Poonam LigadePoonam Ligade
413
$endgroup$
You will get better results if you change subset of images from downsampled class randomly at each epoch. It will help in building a robust model as well. And use data augmentation techniques for minor classes.
answered Sep 26 '18 at 10:24
Poonam LigadePoonam Ligade
413
edited Sep 26 '18 at 11:24
answered Sep 26 '18 at 10:24
Poonam LigadePoonam Ligade
413
answered Sep 26 '18 at 10:24
Poonam LigadePoonam Ligade
413
answered Sep 26 '18 at 10:24
Poonam LigadePoonam Ligade
413
413
$begingroup$
I guess you missed my question. I am not asking how to balanced my dataset, I am asking should I changed my down-sampled class at each epoch or keep it same during traning of my neural net.
$endgroup$
– zwlayer
Sep 26 '18 at 11:02
add a comment |
$begingroup$
I guess you missed my question. I am not asking how to balanced my dataset, I am asking should I changed my down-sampled class at each epoch or keep it same during traning of my neural net.
$endgroup$
– zwlayer
Sep 26 '18 at 11:02
$begingroup$
I guess you missed my question. I am not asking how to balanced my dataset, I am asking should I changed my down-sampled class at each epoch or keep it same during traning of my neural net.
$endgroup$
– zwlayer
Sep 26 '18 at 11:02
$begingroup$
I guess you missed my question. I am not asking how to balanced my dataset, I am asking should I changed my down-sampled class at each epoch or keep it same during traning of my neural net.
$endgroup$
– zwlayer
Sep 26 '18 at 11:02
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%2f38805%2fdownsampling-the-dataset-to-create-balanced-dataset-for-neural-models%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
