Trjtdtk

Why do we bend a book to keep it straight?

What's the meaning of "fortified infraction restraint"?

Can anything be seen from the center of the Boötes void? How dark would it be?

Are all finite dimensional hilbert spaces isomorphic to spaces with Euclidean norms?

An adverb for when you're not exaggerating

How often does castling occur in grandmaster games?

Is grep documentation wrong?

Do jazz musicians improvise on the parent scale in addition to the chord-scales?

When a candle burns, why does the top of wick glow if bottom of flame is hottest?

Trademark violation for app?

How to convince students of the implication truth values?

Did Deadpool rescue all of the X-Force?

If a VARCHAR(MAX) column is included in an index, is the entire value always stored in the index page(s)?

Is this homebrew Lady of Pain warlock patron balanced?

What causes the direction of lightning flashes?

How would a mousetrap for use in space work?

Time to Settle Down!

Is there any word for a place full of confusion?

why is Nikon 1.4g better when Nikon 1.8g is sharper?

Can a new player join a group only when a new campaign starts?

What is the meaning of the simile “quick as silk”?

If windows 7 doesn't support WSL, then what does Linux subsystem option mean?

As a beginner, should I get a Squier Strat with a SSS config or a HSS?

If my PI received research grants from a company to be able to pay my postdoc salary, did I have a potential conflict interest too?

Why margin-based ranking loss is reversed in these two papers?

1

$begingroup$

For knwoledge graph completion, it is very common to use margin-based ranking loss

In the paper:margin-based ranking loss is defined as

$$ min sum_(h,l,t)in S sum_(h',l,t')in S'[gamma + d(h,l,t) - d(h',l,t')]_+$$

Here $d(cdot)$ is the predictive model, $(h,l,t)$ means a positive training instance, and $(h',l,t')$ means a negative training instance corresponding to $(h,l,t)$.

However, in the Andrew's paper, it defines

$$ min sum_(h,l,t)in S sum_(h',l,t')in S'[gamma + d(h',l,t') - d(h,l,t)]_+$$

It seems that they switch the terms $d(h',l,t')$ and $d(h,l,t)$.

My question is that

does it matter to switch $d(h',l,t')$ and $d(h,l,t)$? it's real strange definition. Thanks

deep-learning loss-function hinge-loss

share|improve this question

edited Apr 2 at 8:34

Esmailian

3,391420

asked Apr 1 at 14:34

jasonjason

1444

$endgroup$

add a comment | 

1

$begingroup$

For knwoledge graph completion, it is very common to use margin-based ranking loss

In the paper:margin-based ranking loss is defined as

$$ min sum_(h,l,t)in S sum_(h',l,t')in S'[gamma + d(h,l,t) - d(h',l,t')]_+$$

Here $d(cdot)$ is the predictive model, $(h,l,t)$ means a positive training instance, and $(h',l,t')$ means a negative training instance corresponding to $(h,l,t)$.

However, in the Andrew's paper, it defines

$$ min sum_(h,l,t)in S sum_(h',l,t')in S'[gamma + d(h',l,t') - d(h,l,t)]_+$$

It seems that they switch the terms $d(h',l,t')$ and $d(h,l,t)$.

My question is that

does it matter to switch $d(h',l,t')$ and $d(h,l,t)$? it's real strange definition. Thanks

deep-learning loss-function hinge-loss

share|improve this question

edited Apr 2 at 8:34

Esmailian

3,391420

asked Apr 1 at 14:34

jasonjason

1444

$endgroup$

add a comment | 

$begingroup$

For knwoledge graph completion, it is very common to use margin-based ranking loss

In the paper:margin-based ranking loss is defined as

$$ min sum_(h,l,t)in S sum_(h',l,t')in S'[gamma + d(h,l,t) - d(h',l,t')]_+$$

Here $d(cdot)$ is the predictive model, $(h,l,t)$ means a positive training instance, and $(h',l,t')$ means a negative training instance corresponding to $(h,l,t)$.

However, in the Andrew's paper, it defines

$$ min sum_(h,l,t)in S sum_(h',l,t')in S'[gamma + d(h',l,t') - d(h,l,t)]_+$$

It seems that they switch the terms $d(h',l,t')$ and $d(h,l,t)$.

My question is that

does it matter to switch $d(h',l,t')$ and $d(h,l,t)$? it's real strange definition. Thanks

deep-learning loss-function hinge-loss

share|improve this question

edited Apr 2 at 8:34

Esmailian

3,391420

asked Apr 1 at 14:34

jasonjason

1444

$endgroup$

share|improve this question

edited Apr 2 at 8:34

Esmailian

3,391420

asked Apr 1 at 14:34

jasonjason

1444

share|improve this question

edited Apr 2 at 8:34

Esmailian

3,391420

asked Apr 1 at 14:34

jasonjason

1444

edited Apr 2 at 8:34

Esmailian

3,391420

edited Apr 2 at 8:34

Esmailian

3,391420

asked Apr 1 at 14:34

jasonjason

1444

asked Apr 1 at 14:34

jasonjason

1444

1 Answer
1

active

oldest

votes

2

$begingroup$

In this paper, $d$ denotes "dissimilarity" which should be minimized for positive samples.

In this paper, $d$ ($g$ in the paper) denotes "similarity" which should be maximized for positive samples (or equivalently $-gleft(T^(i)right)$ should be minimized)

share|improve this answer

answered Apr 1 at 14:47

EsmailianEsmailian

3,391420

$endgroup$

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
);



);

draft saved

draft discarded

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

Name

Email

Required, but never shown

StackExchange.ready(
function ()
StackExchange.openid.initPostLogin('.new-post-login', 'https%3a%2f%2fdatascience.stackexchange.com%2fquestions%2f48357%2fwhy-margin-based-ranking-loss-is-reversed-in-these-two-papers%23new-answer', 'question_page');

);

Post as a guest

Name

Email

Required, but never shown

2

$begingroup$

In this paper, $d$ denotes "dissimilarity" which should be minimized for positive samples.

In this paper, $d$ ($g$ in the paper) denotes "similarity" which should be maximized for positive samples (or equivalently $-gleft(T^(i)right)$ should be minimized)

share|improve this answer

answered Apr 1 at 14:47

EsmailianEsmailian

3,391420

$endgroup$

add a comment | 

2

$begingroup$

In this paper, $d$ denotes "dissimilarity" which should be minimized for positive samples.

In this paper, $d$ ($g$ in the paper) denotes "similarity" which should be maximized for positive samples (or equivalently $-gleft(T^(i)right)$ should be minimized)

share|improve this answer

answered Apr 1 at 14:47

EsmailianEsmailian

3,391420

$endgroup$

add a comment | 

$begingroup$

In this paper, $d$ denotes "dissimilarity" which should be minimized for positive samples.

In this paper, $d$ ($g$ in the paper) denotes "similarity" which should be maximized for positive samples (or equivalently $-gleft(T^(i)right)$ should be minimized)

share|improve this answer

answered Apr 1 at 14:47

EsmailianEsmailian

3,391420

$endgroup$

share|improve this answer

answered Apr 1 at 14:47

EsmailianEsmailian

3,391420

answered Apr 1 at 14:47

EsmailianEsmailian

3,391420

answered Apr 1 at 14:47

EsmailianEsmailian

3,391420