Metric Learning vs Similarity LearningWhat is the best Battleship AI?What is the difference between supervised learning and unsupervised learning?NLP and Machine learning for sentiment analysisHow do you measure similarity between 2 series of data?Find the similarity metric between two stringsWhat is difference between java measure frameworks like Metrics and Apache sirona?metric learning for information retrieval in semi-structured text?Dropwizard Metric aggregation issues on graphiteML - Instance-based learningLearning metric/signal similarity using embeddings/autoencoder
Do Rabbis get punished in Heaven for wrong interpretations or claims?
Would it be a good idea to memorize relative interval positions on guitar?
"I you already know": is this proper English?
Why did Saturn V not head straight to the moon?
Why is chess failing to attract big name sponsors?
Are glider winch launches rarer in the USA than in the rest of the world? Why?
How may I shorten this shell script?
Which creatures count as green creatures?
Big Sample size, Small coefficients, significant results. What should I do?
High income, sudden windfall
What do I do when a student working in my lab "ghosts" me?
kids pooling money for Lego League and taxes
Sextortion with actual password not found in leaks
How do I run a game when my PCs have different approaches to combat?
Should I describe a character deeply before killing it?
What do teaching faculty do during semester breaks?
Character Frequency in a String
Why did NASA use U.S customary units?
USA: Can a witness take the 5th to avoid perjury?
Book with a female main character living in a convent who has to fight gods
Does static fire reduce reliability?
How can I stop myself from micromanaging other PCs' actions?
How important is a good quality camera for good photography?
Who has jurisdiction for a crime committed in an embassy?
Metric Learning vs Similarity Learning
What is the best Battleship AI?What is the difference between supervised learning and unsupervised learning?NLP and Machine learning for sentiment analysisHow do you measure similarity between 2 series of data?Find the similarity metric between two stringsWhat is difference between java measure frameworks like Metrics and Apache sirona?metric learning for information retrieval in semi-structured text?Dropwizard Metric aggregation issues on graphiteML - Instance-based learningLearning metric/signal similarity using embeddings/autoencoder
.everyoneloves__top-leaderboard:empty,.everyoneloves__mid-leaderboard:empty,.everyoneloves__bot-mid-leaderboard:empty margin-bottom:0;
I struggle to find any comprehensive explanations about Similarity Learning. From what I have gathered it is the same as Metric learning, except it attempts to learn a similarity function rather than a metric.
Can anyone please clarify the difference between them?
Any links or sources would be greatly appreciated.
Thanks in advance.
machine-learning artificial-intelligence metrics similarity ranking
asked Mar 26 at 15:50
Rabbo PaschRabbo Pasch
434 bronze badges
add a comment |
I struggle to find any comprehensive explanations about Similarity Learning. From what I have gathered it is the same as Metric learning, except it attempts to learn a similarity function rather than a metric.
Can anyone please clarify the difference between them?
Any links or sources would be greatly appreciated.
Thanks in advance.
machine-learning artificial-intelligence metrics similarity ranking
asked Mar 26 at 15:50
Rabbo PaschRabbo Pasch
434 bronze badges
What have you found that is comprehensible but not sufficiently comprehensive? The Wikipedia article covers it quite well.
– Prune
Mar 26 at 22:47
So I understand that both attempt to learn a metric or similarity function that can relate or distinguished two objects. I however dont understand the difference in the concepts of metric and similarity. It feels like they are almost the same measure.
– Rabbo Pasch
Mar 28 at 6:23
add a comment |
I struggle to find any comprehensive explanations about Similarity Learning. From what I have gathered it is the same as Metric learning, except it attempts to learn a similarity function rather than a metric.
Can anyone please clarify the difference between them?
Any links or sources would be greatly appreciated.
Thanks in advance.
machine-learning artificial-intelligence metrics similarity ranking
asked Mar 26 at 15:50
Rabbo PaschRabbo Pasch
434 bronze badges
I struggle to find any comprehensive explanations about Similarity Learning. From what I have gathered it is the same as Metric learning, except it attempts to learn a similarity function rather than a metric.
Can anyone please clarify the difference between them?
Any links or sources would be greatly appreciated.
Thanks in advance.
machine-learning artificial-intelligence metrics similarity ranking
machine-learning artificial-intelligence metrics similarity ranking
asked Mar 26 at 15:50
Rabbo PaschRabbo Pasch
434 bronze badges
asked Mar 26 at 15:50
Rabbo PaschRabbo Pasch
434 bronze badges
asked Mar 26 at 15:50
Rabbo PaschRabbo Pasch
434 bronze badges
asked Mar 26 at 15:50
Rabbo PaschRabbo Pasch
434 bronze badges
asked Mar 26 at 15:50
Rabbo PaschRabbo Pasch
434 bronze badges
434 bronze badges
What have you found that is comprehensible but not sufficiently comprehensive? The Wikipedia article covers it quite well.
– Prune
Mar 26 at 22:47
So I understand that both attempt to learn a metric or similarity function that can relate or distinguished two objects. I however dont understand the difference in the concepts of metric and similarity. It feels like they are almost the same measure.
– Rabbo Pasch
Mar 28 at 6:23
add a comment |
What have you found that is comprehensible but not sufficiently comprehensive? The Wikipedia article covers it quite well.
– Prune
Mar 26 at 22:47
So I understand that both attempt to learn a metric or similarity function that can relate or distinguished two objects. I however dont understand the difference in the concepts of metric and similarity. It feels like they are almost the same measure.
– Rabbo Pasch
Mar 28 at 6:23
What have you found that is comprehensible but not sufficiently comprehensive? The Wikipedia article covers it quite well.
– Prune
Mar 26 at 22:47
What have you found that is comprehensible but not sufficiently comprehensive? The Wikipedia article covers it quite well.
– Prune
Mar 26 at 22:47
So I understand that both attempt to learn a metric or similarity function that can relate or distinguished two objects. I however dont understand the difference in the concepts of metric and similarity. It feels like they are almost the same measure.
– Rabbo Pasch
Mar 28 at 6:23
So I understand that both attempt to learn a metric or similarity function that can relate or distinguished two objects. I however dont understand the difference in the concepts of metric and similarity. It feels like they are almost the same measure.
– Rabbo Pasch
Mar 28 at 6:23
add a comment |
1 Answer
1
active
oldest
votes
For most (all?) purposes, metric learning is a subset of similarity learning. Note that, in common use, "similar" is roughly an inverse of "distance": things with a low distance between them have high similarity. In practice, this is usually a matter of semantic choice -- a continuous transformation can generally make the two isomorphic.
A metric needs to follow certain rules; a similarity function has looser standards. For instance, compare a full-length (say, 2 hours) movie M with a 20-minute animated reduction A. A metric function f requires that f(M, A) = f(A, M). However, if you decide that the richness of the movie means that it shouldn't regard the cartoon as such close kin, you might input the pair of training triples
(A, M, 0.90)
(M, A, 0.15)
Another example would be with set similarity, measured by size and membership, but in a non-Euclidean fashion.
a = 1, 2, 3, 4
b = 3, 4, 5, 6
c = 5, 6, 7, 8
A similarity training would allow
(a, b, 2)
(b, c, 2)
(a, c, 10)
In this "world", a and c suffer a large penalty because they have nothing in common but set size. b is close to each of them due to having half the elements in common. This would give a metric function a headache, since it severely violates subaddition, the triangle inequality.
Does that help clear up the differences?
answered Mar 26 at 23:11
PrunePrune
50.1k14 gold badges38 silver badges61 bronze badges
This does clear things up. Is it correct if I say that a Similarity measure is a more abstract distance metric that can not be represented in a Euclidean space?
– Rabbo Pasch
Mar 28 at 6:28
No: a similarity measure does not have to qualify as a distance metric. "distance metric" is a hard-lined defined term. That would be somewhat like saying that a general complex number is a "more abstract integer".
– Prune
Mar 28 at 16:10
Yes, that is how I understood it, my wording might have been a little off regarding 'Similarity is a more abstract..' Thank you
– Rabbo Pasch
Mar 29 at 19:40
I saw that you are a editor of research papers and a deep learning engineer. I am currently working on a research paper on using deep similarity learning to predict football match outcomes and their rankings. Just thought that you might be interested in the topic and the final product. If you are, let me know.
– Rabbo Pasch
Mar 29 at 19:45
Good to know that we're on the same thought track; thanks. Yes, I'd be interested in the research paper. Post a link here when the time comes?
– Prune
Mar 29 at 20:30
add a comment |
Your Answer
StackExchange.ifUsing("editor", function ()
StackExchange.using("externalEditor", function ()
StackExchange.using("snippets", function ()
StackExchange.snippets.init();
);
);
, "code-snippets");
StackExchange.ready(function()
var channelOptions =
tags: "".split(" "),
id: "1"
;
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
,
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%2fstackoverflow.com%2fquestions%2f55361279%2fmetric-learning-vs-similarity-learning%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
For most (all?) purposes, metric learning is a subset of similarity learning. Note that, in common use, "similar" is roughly an inverse of "distance": things with a low distance between them have high similarity. In practice, this is usually a matter of semantic choice -- a continuous transformation can generally make the two isomorphic.
A metric needs to follow certain rules; a similarity function has looser standards. For instance, compare a full-length (say, 2 hours) movie M with a 20-minute animated reduction A. A metric function f requires that f(M, A) = f(A, M). However, if you decide that the richness of the movie means that it shouldn't regard the cartoon as such close kin, you might input the pair of training triples
(A, M, 0.90)
(M, A, 0.15)
Another example would be with set similarity, measured by size and membership, but in a non-Euclidean fashion.
a = 1, 2, 3, 4
b = 3, 4, 5, 6
c = 5, 6, 7, 8
A similarity training would allow
(a, b, 2)
(b, c, 2)
(a, c, 10)
In this "world", a and c suffer a large penalty because they have nothing in common but set size. b is close to each of them due to having half the elements in common. This would give a metric function a headache, since it severely violates subaddition, the triangle inequality.
Does that help clear up the differences?
answered Mar 26 at 23:11
PrunePrune
50.1k14 gold badges38 silver badges61 bronze badges
This does clear things up. Is it correct if I say that a Similarity measure is a more abstract distance metric that can not be represented in a Euclidean space?
– Rabbo Pasch
Mar 28 at 6:28
No: a similarity measure does not have to qualify as a distance metric. "distance metric" is a hard-lined defined term. That would be somewhat like saying that a general complex number is a "more abstract integer".
– Prune
Mar 28 at 16:10
Yes, that is how I understood it, my wording might have been a little off regarding 'Similarity is a more abstract..' Thank you
– Rabbo Pasch
Mar 29 at 19:40
I saw that you are a editor of research papers and a deep learning engineer. I am currently working on a research paper on using deep similarity learning to predict football match outcomes and their rankings. Just thought that you might be interested in the topic and the final product. If you are, let me know.
– Rabbo Pasch
Mar 29 at 19:45
Good to know that we're on the same thought track; thanks. Yes, I'd be interested in the research paper. Post a link here when the time comes?
– Prune
Mar 29 at 20:30
add a comment |
For most (all?) purposes, metric learning is a subset of similarity learning. Note that, in common use, "similar" is roughly an inverse of "distance": things with a low distance between them have high similarity. In practice, this is usually a matter of semantic choice -- a continuous transformation can generally make the two isomorphic.
A metric needs to follow certain rules; a similarity function has looser standards. For instance, compare a full-length (say, 2 hours) movie M with a 20-minute animated reduction A. A metric function f requires that f(M, A) = f(A, M). However, if you decide that the richness of the movie means that it shouldn't regard the cartoon as such close kin, you might input the pair of training triples
(A, M, 0.90)
(M, A, 0.15)
Another example would be with set similarity, measured by size and membership, but in a non-Euclidean fashion.
a = 1, 2, 3, 4
b = 3, 4, 5, 6
c = 5, 6, 7, 8
A similarity training would allow
(a, b, 2)
(b, c, 2)
(a, c, 10)
In this "world", a and c suffer a large penalty because they have nothing in common but set size. b is close to each of them due to having half the elements in common. This would give a metric function a headache, since it severely violates subaddition, the triangle inequality.
Does that help clear up the differences?
answered Mar 26 at 23:11
PrunePrune
50.1k14 gold badges38 silver badges61 bronze badges
This does clear things up. Is it correct if I say that a Similarity measure is a more abstract distance metric that can not be represented in a Euclidean space?
– Rabbo Pasch
Mar 28 at 6:28
No: a similarity measure does not have to qualify as a distance metric. "distance metric" is a hard-lined defined term. That would be somewhat like saying that a general complex number is a "more abstract integer".
– Prune
Mar 28 at 16:10
Yes, that is how I understood it, my wording might have been a little off regarding 'Similarity is a more abstract..' Thank you
– Rabbo Pasch
Mar 29 at 19:40
I saw that you are a editor of research papers and a deep learning engineer. I am currently working on a research paper on using deep similarity learning to predict football match outcomes and their rankings. Just thought that you might be interested in the topic and the final product. If you are, let me know.
– Rabbo Pasch
Mar 29 at 19:45
Good to know that we're on the same thought track; thanks. Yes, I'd be interested in the research paper. Post a link here when the time comes?
– Prune
Mar 29 at 20:30
add a comment |
For most (all?) purposes, metric learning is a subset of similarity learning. Note that, in common use, "similar" is roughly an inverse of "distance": things with a low distance between them have high similarity. In practice, this is usually a matter of semantic choice -- a continuous transformation can generally make the two isomorphic.
A metric needs to follow certain rules; a similarity function has looser standards. For instance, compare a full-length (say, 2 hours) movie M with a 20-minute animated reduction A. A metric function f requires that f(M, A) = f(A, M). However, if you decide that the richness of the movie means that it shouldn't regard the cartoon as such close kin, you might input the pair of training triples
(A, M, 0.90)
(M, A, 0.15)
Another example would be with set similarity, measured by size and membership, but in a non-Euclidean fashion.
a = 1, 2, 3, 4
b = 3, 4, 5, 6
c = 5, 6, 7, 8
A similarity training would allow
(a, b, 2)
(b, c, 2)
(a, c, 10)
In this "world", a and c suffer a large penalty because they have nothing in common but set size. b is close to each of them due to having half the elements in common. This would give a metric function a headache, since it severely violates subaddition, the triangle inequality.
Does that help clear up the differences?
answered Mar 26 at 23:11
PrunePrune
50.1k14 gold badges38 silver badges61 bronze badges
For most (all?) purposes, metric learning is a subset of similarity learning. Note that, in common use, "similar" is roughly an inverse of "distance": things with a low distance between them have high similarity. In practice, this is usually a matter of semantic choice -- a continuous transformation can generally make the two isomorphic.
A metric needs to follow certain rules; a similarity function has looser standards. For instance, compare a full-length (say, 2 hours) movie M with a 20-minute animated reduction A. A metric function f requires that f(M, A) = f(A, M). However, if you decide that the richness of the movie means that it shouldn't regard the cartoon as such close kin, you might input the pair of training triples
(A, M, 0.90)
(M, A, 0.15)
Another example would be with set similarity, measured by size and membership, but in a non-Euclidean fashion.
a = 1, 2, 3, 4
b = 3, 4, 5, 6
c = 5, 6, 7, 8
A similarity training would allow
(a, b, 2)
(b, c, 2)
(a, c, 10)
In this "world", a and c suffer a large penalty because they have nothing in common but set size. b is close to each of them due to having half the elements in common. This would give a metric function a headache, since it severely violates subaddition, the triangle inequality.
Does that help clear up the differences?
answered Mar 26 at 23:11
PrunePrune
50.1k14 gold badges38 silver badges61 bronze badges
answered Mar 26 at 23:11
PrunePrune
50.1k14 gold badges38 silver badges61 bronze badges
answered Mar 26 at 23:11
PrunePrune
50.1k14 gold badges38 silver badges61 bronze badges
answered Mar 26 at 23:11
PrunePrune
50.1k14 gold badges38 silver badges61 bronze badges
50.1k14 gold badges38 silver badges61 bronze badges
This does clear things up. Is it correct if I say that a Similarity measure is a more abstract distance metric that can not be represented in a Euclidean space?
– Rabbo Pasch
Mar 28 at 6:28
No: a similarity measure does not have to qualify as a distance metric. "distance metric" is a hard-lined defined term. That would be somewhat like saying that a general complex number is a "more abstract integer".
– Prune
Mar 28 at 16:10
Yes, that is how I understood it, my wording might have been a little off regarding 'Similarity is a more abstract..' Thank you
– Rabbo Pasch
Mar 29 at 19:40
I saw that you are a editor of research papers and a deep learning engineer. I am currently working on a research paper on using deep similarity learning to predict football match outcomes and their rankings. Just thought that you might be interested in the topic and the final product. If you are, let me know.
– Rabbo Pasch
Mar 29 at 19:45
Good to know that we're on the same thought track; thanks. Yes, I'd be interested in the research paper. Post a link here when the time comes?
– Prune
Mar 29 at 20:30
add a comment |
This does clear things up. Is it correct if I say that a Similarity measure is a more abstract distance metric that can not be represented in a Euclidean space?
– Rabbo Pasch
Mar 28 at 6:28
No: a similarity measure does not have to qualify as a distance metric. "distance metric" is a hard-lined defined term. That would be somewhat like saying that a general complex number is a "more abstract integer".
– Prune
Mar 28 at 16:10
Yes, that is how I understood it, my wording might have been a little off regarding 'Similarity is a more abstract..' Thank you
– Rabbo Pasch
Mar 29 at 19:40
I saw that you are a editor of research papers and a deep learning engineer. I am currently working on a research paper on using deep similarity learning to predict football match outcomes and their rankings. Just thought that you might be interested in the topic and the final product. If you are, let me know.
– Rabbo Pasch
Mar 29 at 19:45
Good to know that we're on the same thought track; thanks. Yes, I'd be interested in the research paper. Post a link here when the time comes?
– Prune
Mar 29 at 20:30
This does clear things up. Is it correct if I say that a Similarity measure is a more abstract distance metric that can not be represented in a Euclidean space?
– Rabbo Pasch
Mar 28 at 6:28
This does clear things up. Is it correct if I say that a Similarity measure is a more abstract distance metric that can not be represented in a Euclidean space?
– Rabbo Pasch
Mar 28 at 6:28
No: a similarity measure does not have to qualify as a distance metric. "distance metric" is a hard-lined defined term. That would be somewhat like saying that a general complex number is a "more abstract integer".
– Prune
Mar 28 at 16:10
No: a similarity measure does not have to qualify as a distance metric. "distance metric" is a hard-lined defined term. That would be somewhat like saying that a general complex number is a "more abstract integer".
– Prune
Mar 28 at 16:10
Yes, that is how I understood it, my wording might have been a little off regarding 'Similarity is a more abstract..' Thank you
– Rabbo Pasch
Mar 29 at 19:40
Yes, that is how I understood it, my wording might have been a little off regarding 'Similarity is a more abstract..' Thank you
– Rabbo Pasch
Mar 29 at 19:40
I saw that you are a editor of research papers and a deep learning engineer. I am currently working on a research paper on using deep similarity learning to predict football match outcomes and their rankings. Just thought that you might be interested in the topic and the final product. If you are, let me know.
– Rabbo Pasch
Mar 29 at 19:45
I saw that you are a editor of research papers and a deep learning engineer. I am currently working on a research paper on using deep similarity learning to predict football match outcomes and their rankings. Just thought that you might be interested in the topic and the final product. If you are, let me know.
– Rabbo Pasch
Mar 29 at 19:45
Good to know that we're on the same thought track; thanks. Yes, I'd be interested in the research paper. Post a link here when the time comes?
– Prune
Mar 29 at 20:30
Good to know that we're on the same thought track; thanks. Yes, I'd be interested in the research paper. Post a link here when the time comes?
– Prune
Mar 29 at 20:30
add a comment |
Got a question that you can’t ask on public Stack Overflow? Learn more about sharing private information with Stack Overflow for Teams.
Got a question that you can’t ask on public Stack Overflow? Learn more about sharing private information with Stack Overflow for Teams.
Thanks for contributing an answer to Stack Overflow!
- 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.
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%2fstackoverflow.com%2fquestions%2f55361279%2fmetric-learning-vs-similarity-learning%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
What have you found that is comprehensible but not sufficiently comprehensive? The Wikipedia article covers it quite well.
– Prune
Mar 26 at 22:47
So I understand that both attempt to learn a metric or similarity function that can relate or distinguished two objects. I however dont understand the difference in the concepts of metric and similarity. It feels like they are almost the same measure.
– Rabbo Pasch
Mar 28 at 6:23