Hfrxdjt

Why is there so little support for joining EFTA in the British parliament?

Did John Wesley plagiarize Matthew Henry...?

Dinosaur Word Search, Letter Solve, and Unscramble

Searching extreme points of polyhedron

Why not use the yoke to control yaw, as well as pitch and roll?

Understanding piped commands in GNU/Linux

Why do C and C++ allow the expression (int) + 4*5?

How to make an animal which can only breed for a certain number of generations?

How can I prevent/balance waiting and turtling as a response to cooldown mechanics

How does the body cool itself in a stillsuit?

How do you write "wild blueberries flavored"?

How many time has Arya actually used Needle?

Weaponising the Grasp-at-a-Distance spell

Should man-made satellites feature an intelligent inverted "cow catcher"?

How to resize main filesystem

What is "Lambda" in Heston's original paper on stochastic volatility models?

Does the universe have a fixed centre of mass?

Is there night in Alpha Complex?

Keep at all times, the minus sign above aligned with minus sign below

Why can't fire hurt Daenerys but it did to Jon Snow in season 1?

Why do the Z-fighters hide their power?

How to get a flat-head nail out of a piece of wood?

NIntegrate on a solution of a matrix ODE

Where did Ptolemy compare the Earth to the distance of fixed stars?

focal loss function help

1

$begingroup$

I am working on a relation extraction and classification problem. The data is in the form of text files. The data is imbalanced. I want to use focal loss function to address class imbalance problem in the data. My question is: Can focal loss be utilized for extraction and classification task to increase the accuracy? Focal loss has been applied on object detection task and for image classification task. The link is below. I want to use this on text classification task.

https://shaoanlu.wordpress.com/2017/08/16/applying-focal-loss-on-cats-vs-dogs-classification-task/

http://openaccess.thecvf.com/content_ICCV_2017/papers/Lin_Focal_Loss_for_ICCV_2017_paper.pdf

multiclass-classification supervised-learning loss-function class-imbalance

share|improve this question

edited Jul 21 '18 at 18:50

M. Ahmad

asked Jul 20 '18 at 20:19

M. AhmadM. Ahmad

63

$endgroup$

bumped to the homepage by Community♦ 28 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 | 

1

$begingroup$

I am working on a relation extraction and classification problem. The data is in the form of text files. The data is imbalanced. I want to use focal loss function to address class imbalance problem in the data. My question is: Can focal loss be utilized for extraction and classification task to increase the accuracy? Focal loss has been applied on object detection task and for image classification task. The link is below. I want to use this on text classification task.

https://shaoanlu.wordpress.com/2017/08/16/applying-focal-loss-on-cats-vs-dogs-classification-task/

http://openaccess.thecvf.com/content_ICCV_2017/papers/Lin_Focal_Loss_for_ICCV_2017_paper.pdf

multiclass-classification supervised-learning loss-function class-imbalance

share|improve this question

edited Jul 21 '18 at 18:50

M. Ahmad

asked Jul 20 '18 at 20:19

M. AhmadM. Ahmad

63

$endgroup$

bumped to the homepage by Community♦ 28 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 am working on a relation extraction and classification problem. The data is in the form of text files. The data is imbalanced. I want to use focal loss function to address class imbalance problem in the data. My question is: Can focal loss be utilized for extraction and classification task to increase the accuracy? Focal loss has been applied on object detection task and for image classification task. The link is below. I want to use this on text classification task.

https://shaoanlu.wordpress.com/2017/08/16/applying-focal-loss-on-cats-vs-dogs-classification-task/

http://openaccess.thecvf.com/content_ICCV_2017/papers/Lin_Focal_Loss_for_ICCV_2017_paper.pdf

multiclass-classification supervised-learning loss-function class-imbalance

share|improve this question

edited Jul 21 '18 at 18:50

M. Ahmad

asked Jul 20 '18 at 20:19

M. AhmadM. Ahmad

63

$endgroup$

share|improve this question

edited Jul 21 '18 at 18:50

M. Ahmad

asked Jul 20 '18 at 20:19

M. AhmadM. Ahmad

63

share|improve this question

edited Jul 21 '18 at 18:50

M. Ahmad

asked Jul 20 '18 at 20:19

M. AhmadM. Ahmad

63

asked Jul 20 '18 at 20:19

M. AhmadM. Ahmad

63

asked Jul 20 '18 at 20:19

M. AhmadM. Ahmad

63

1 Answer
1

active

oldest

votes

0

$begingroup$

From Focal Loss for Dense Object Detection paper, focal loss is defined as:

$$FL(p_t) = −(1 − p_t)^γ log(p_t)$$

It is a generalized formation of a loss function that can be directly implemented for any classification problem, including text classification.

Focal loss function is designed for problems with 1:1000 training imbalance. Does your data actually have that much imbalance? If not, you can use much simpler methods (i.e., resampling).

share|improve this answer

edited Jul 23 '18 at 0:36

answered Jul 21 '18 at 21:01

Brian SpieringBrian Spiering

4,2881129

$endgroup$

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%2f35813%2ffocal-loss-function-help%23new-answer', 'question_page');

);

Post as a guest

Name

Email

Required, but never shown

0

$begingroup$

From Focal Loss for Dense Object Detection paper, focal loss is defined as:

$$FL(p_t) = −(1 − p_t)^γ log(p_t)$$

It is a generalized formation of a loss function that can be directly implemented for any classification problem, including text classification.

Focal loss function is designed for problems with 1:1000 training imbalance. Does your data actually have that much imbalance? If not, you can use much simpler methods (i.e., resampling).

share|improve this answer

edited Jul 23 '18 at 0:36

answered Jul 21 '18 at 21:01

Brian SpieringBrian Spiering

4,2881129

$endgroup$

add a comment | 

0

$begingroup$

From Focal Loss for Dense Object Detection paper, focal loss is defined as:

$$FL(p_t) = −(1 − p_t)^γ log(p_t)$$

It is a generalized formation of a loss function that can be directly implemented for any classification problem, including text classification.

Focal loss function is designed for problems with 1:1000 training imbalance. Does your data actually have that much imbalance? If not, you can use much simpler methods (i.e., resampling).

share|improve this answer

edited Jul 23 '18 at 0:36

answered Jul 21 '18 at 21:01

Brian SpieringBrian Spiering

4,2881129

$endgroup$

add a comment | 

$begingroup$

From Focal Loss for Dense Object Detection paper, focal loss is defined as:

$$FL(p_t) = −(1 − p_t)^γ log(p_t)$$

It is a generalized formation of a loss function that can be directly implemented for any classification problem, including text classification.

Focal loss function is designed for problems with 1:1000 training imbalance. Does your data actually have that much imbalance? If not, you can use much simpler methods (i.e., resampling).

share|improve this answer

edited Jul 23 '18 at 0:36

answered Jul 21 '18 at 21:01

Brian SpieringBrian Spiering

4,2881129

$endgroup$

share|improve this answer

edited Jul 23 '18 at 0:36

answered Jul 21 '18 at 21:01

Brian SpieringBrian Spiering

4,2881129

answered Jul 21 '18 at 21:01

Brian SpieringBrian Spiering

4,2881129

answered Jul 21 '18 at 21:01

Brian SpieringBrian Spiering

4,2881129