Combining multiple Principal Component Analysis (PCA) feature extractors into one. How to do it?
$begingroup$
I have a set of images (1000) of cats (32x32 pixels ~> flatten:1024pixels ), and want to use Principal Component Analysis to extract some features. Lets say 100 features.
Then I have basis of U_1, ... , U_1024, and i only use the first 100 vectors to extract the features. So a picture of a cat A in R^1024 is mapped to a vector x in R^100, s.th. U*x ~= A. I.e.
This works fine.
Now I get another set of 1000 images, this time dogs. I do the same procedure and get V_1, ... , V_100.
I assume that many features are quite similar, but some are different.
What I want to do now is combine the features U_1, ..., U_100, V_1, ... , V_100 to a new set of feature W_1, ... , W_100 (or maybe W_120 for example), such that I can use W on both cats and dogs. I don't want 200 features because they will be repetitive. But how to combine U's and V's in a smart way?
(Concretely my problem is that I have many sets of 500 images, and to do PCA on all images is not feasible. So I want to break it apart, and combine the PCA's.)
statistics machine-learning
edited Dec 16 '18 at 21:52
Bernard
121k740116
asked Dec 16 '18 at 21:35
spookyspooky
403
$endgroup$
add a comment |
$begingroup$
I have a set of images (1000) of cats (32x32 pixels ~> flatten:1024pixels ), and want to use Principal Component Analysis to extract some features. Lets say 100 features.
Then I have basis of U_1, ... , U_1024, and i only use the first 100 vectors to extract the features. So a picture of a cat A in R^1024 is mapped to a vector x in R^100, s.th. U*x ~= A. I.e.
This works fine.
Now I get another set of 1000 images, this time dogs. I do the same procedure and get V_1, ... , V_100.
I assume that many features are quite similar, but some are different.
What I want to do now is combine the features U_1, ..., U_100, V_1, ... , V_100 to a new set of feature W_1, ... , W_100 (or maybe W_120 for example), such that I can use W on both cats and dogs. I don't want 200 features because they will be repetitive. But how to combine U's and V's in a smart way?
(Concretely my problem is that I have many sets of 500 images, and to do PCA on all images is not feasible. So I want to break it apart, and combine the PCA's.)
statistics machine-learning
edited Dec 16 '18 at 21:52
Bernard
121k740116
asked Dec 16 '18 at 21:35
spookyspooky
403
$endgroup$
$begingroup$
Why not allow the system to select 120 (or any other preferred number of vectors) from the 2000 pictures? I am not sure that direct combination is so simple - what you attempt to do is to MIMIC Machine Learning into Human Learning - in a way you defeat the idea of machine learning that tells us that the 100 vectors from one set can be used to create a new set. Try to combine 500 of cats with 500 of dogs and see what type of vectors you get?
$endgroup$
– Moti
Dec 17 '18 at 2:19
add a comment |
$begingroup$
I have a set of images (1000) of cats (32x32 pixels ~> flatten:1024pixels ), and want to use Principal Component Analysis to extract some features. Lets say 100 features.
Then I have basis of U_1, ... , U_1024, and i only use the first 100 vectors to extract the features. So a picture of a cat A in R^1024 is mapped to a vector x in R^100, s.th. U*x ~= A. I.e.
This works fine.
Now I get another set of 1000 images, this time dogs. I do the same procedure and get V_1, ... , V_100.
I assume that many features are quite similar, but some are different.
What I want to do now is combine the features U_1, ..., U_100, V_1, ... , V_100 to a new set of feature W_1, ... , W_100 (or maybe W_120 for example), such that I can use W on both cats and dogs. I don't want 200 features because they will be repetitive. But how to combine U's and V's in a smart way?
(Concretely my problem is that I have many sets of 500 images, and to do PCA on all images is not feasible. So I want to break it apart, and combine the PCA's.)
statistics machine-learning
edited Dec 16 '18 at 21:52
Bernard
121k740116
asked Dec 16 '18 at 21:35
spookyspooky
403
$endgroup$
I have a set of images (1000) of cats (32x32 pixels ~> flatten:1024pixels ), and want to use Principal Component Analysis to extract some features. Lets say 100 features.
Then I have basis of U_1, ... , U_1024, and i only use the first 100 vectors to extract the features. So a picture of a cat A in R^1024 is mapped to a vector x in R^100, s.th. U*x ~= A. I.e.
This works fine.
Now I get another set of 1000 images, this time dogs. I do the same procedure and get V_1, ... , V_100.
I assume that many features are quite similar, but some are different.
What I want to do now is combine the features U_1, ..., U_100, V_1, ... , V_100 to a new set of feature W_1, ... , W_100 (or maybe W_120 for example), such that I can use W on both cats and dogs. I don't want 200 features because they will be repetitive. But how to combine U's and V's in a smart way?
(Concretely my problem is that I have many sets of 500 images, and to do PCA on all images is not feasible. So I want to break it apart, and combine the PCA's.)
statistics machine-learning
statistics machine-learning
edited Dec 16 '18 at 21:52
Bernard
121k740116
asked Dec 16 '18 at 21:35
spookyspooky
403
edited Dec 16 '18 at 21:52
Bernard
121k740116
asked Dec 16 '18 at 21:35
spookyspooky
403
edited Dec 16 '18 at 21:52
Bernard
121k740116
edited Dec 16 '18 at 21:52
Bernard
121k740116
edited Dec 16 '18 at 21:52
Bernard
121k740116
121k740116
asked Dec 16 '18 at 21:35
spookyspooky
403
asked Dec 16 '18 at 21:35
spookyspooky
403
asked Dec 16 '18 at 21:35
spookyspooky
403
403
$begingroup$
Why not allow the system to select 120 (or any other preferred number of vectors) from the 2000 pictures? I am not sure that direct combination is so simple - what you attempt to do is to MIMIC Machine Learning into Human Learning - in a way you defeat the idea of machine learning that tells us that the 100 vectors from one set can be used to create a new set. Try to combine 500 of cats with 500 of dogs and see what type of vectors you get?
$endgroup$
– Moti
Dec 17 '18 at 2:19
add a comment |
$begingroup$
Why not allow the system to select 120 (or any other preferred number of vectors) from the 2000 pictures? I am not sure that direct combination is so simple - what you attempt to do is to MIMIC Machine Learning into Human Learning - in a way you defeat the idea of machine learning that tells us that the 100 vectors from one set can be used to create a new set. Try to combine 500 of cats with 500 of dogs and see what type of vectors you get?
$endgroup$
– Moti
Dec 17 '18 at 2:19
$begingroup$
Why not allow the system to select 120 (or any other preferred number of vectors) from the 2000 pictures? I am not sure that direct combination is so simple - what you attempt to do is to MIMIC Machine Learning into Human Learning - in a way you defeat the idea of machine learning that tells us that the 100 vectors from one set can be used to create a new set. Try to combine 500 of cats with 500 of dogs and see what type of vectors you get?
$endgroup$
– Moti
Dec 17 '18 at 2:19
$begingroup$
Why not allow the system to select 120 (or any other preferred number of vectors) from the 2000 pictures? I am not sure that direct combination is so simple - what you attempt to do is to MIMIC Machine Learning into Human Learning - in a way you defeat the idea of machine learning that tells us that the 100 vectors from one set can be used to create a new set. Try to combine 500 of cats with 500 of dogs and see what type of vectors you get?
$endgroup$
– Moti
Dec 17 '18 at 2:19
add a comment |
0
active
oldest
votes
Your Answer
StackExchange.ifUsing("editor", function () {
return StackExchange.using("mathjaxEditing", function () {
StackExchange.MarkdownEditor.creationCallbacks.add(function (editor, postfix) {
StackExchange.mathjaxEditing.prepareWmdForMathJax(editor, postfix, [["$", "$"], ["\\(","\\)"]]);
});
});
}, "mathjax-editing");
StackExchange.ready(function() {
var channelOptions = {
tags: "".split(" "),
id: "69"
};
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: true,
noModals: true,
showLowRepImageUploadWarning: true,
reputationToPostImages: 10,
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
},
noCode: 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%2fmath.stackexchange.com%2fquestions%2f3043199%2fcombining-multiple-principal-component-analysis-pca-feature-extractors-into-on%23new-answer', 'question_page');
}
);
Post as a guest
Required, but never shown
0
active
oldest
votes
0
active
oldest
votes
active
oldest
votes
active
oldest
votes
Thanks for contributing an answer to Mathematics 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%2fmath.stackexchange.com%2fquestions%2f3043199%2fcombining-multiple-principal-component-analysis-pca-feature-extractors-into-on%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
$begingroup$
Why not allow the system to select 120 (or any other preferred number of vectors) from the 2000 pictures? I am not sure that direct combination is so simple - what you attempt to do is to MIMIC Machine Learning into Human Learning - in a way you defeat the idea of machine learning that tells us that the 100 vectors from one set can be used to create a new set. Try to combine 500 of cats with 500 of dogs and see what type of vectors you get?
$endgroup$
– Moti
Dec 17 '18 at 2:19