Existence of MLE












1












$begingroup$


I have a problem with MLE's definition:



Casella Berger in Statistical Inference and Nitis Mukhopadhyay in Probability and Statistics said that MLE for a parameter $thetainTheta$ is respectively $argsup_{thetainTheta}{L(thetamid x)}$ or $argmax_{thetainTheta}{L(thetamid x)}$.



But this estimator is a supremum or a maximum? If it is a supremum why we don't call it supremum likelihood estimator? Conversely if it is a maximum, the likelihood function must be continuous and the parametric space must be compact (sufficient condition for the existence of maximum)



What is the truth or the minimal condition for the existence of MLE?






probability statistics statistical-inference estimation maximum-likelihood







share|cite|improve this question











$endgroup$








  • 2




    $begingroup$
    Do you understand the difference between a "supremum" and a "maximum"? Any set of real numbers, having an upper bound, as a supremum but not necessarily a maximum. A set of real numbers has a maximum if and only if the supremum is in the set in which case the maximum is the supremum.
    $endgroup$
    – user247327
    Jan 2 at 17:47










  • $begingroup$
    Of course I know the difference. My problem is that in an estimation, the estimate parameter must belong to the parametric space but if the parametric space is for example an open set, the supremum may stay in its boundary and this thing may has no sense. So I want to understand which is the correct definition of MLE; there is argsup or argmax? because the second one may not belong to the parametric space. Thanks
    $endgroup$
    – Elia
    Jan 2 at 18:54






  • 1




    $begingroup$
    Then you realize that if the maximum [b]exists[/b], there is no difference between the two. If the maximum does not exist, then you [b]have[/b] to use the supremum.
    $endgroup$
    – user247327
    Jan 3 at 18:12
















1












$begingroup$


I have a problem with MLE's definition:



Casella Berger in Statistical Inference and Nitis Mukhopadhyay in Probability and Statistics said that MLE for a parameter $thetainTheta$ is respectively $argsup_{thetainTheta}{L(thetamid x)}$ or $argmax_{thetainTheta}{L(thetamid x)}$.



But this estimator is a supremum or a maximum? If it is a supremum why we don't call it supremum likelihood estimator? Conversely if it is a maximum, the likelihood function must be continuous and the parametric space must be compact (sufficient condition for the existence of maximum)



What is the truth or the minimal condition for the existence of MLE?






probability statistics statistical-inference estimation maximum-likelihood







share|cite|improve this question











$endgroup$








  • 2




    $begingroup$
    Do you understand the difference between a "supremum" and a "maximum"? Any set of real numbers, having an upper bound, as a supremum but not necessarily a maximum. A set of real numbers has a maximum if and only if the supremum is in the set in which case the maximum is the supremum.
    $endgroup$
    – user247327
    Jan 2 at 17:47










  • $begingroup$
    Of course I know the difference. My problem is that in an estimation, the estimate parameter must belong to the parametric space but if the parametric space is for example an open set, the supremum may stay in its boundary and this thing may has no sense. So I want to understand which is the correct definition of MLE; there is argsup or argmax? because the second one may not belong to the parametric space. Thanks
    $endgroup$
    – Elia
    Jan 2 at 18:54






  • 1




    $begingroup$
    Then you realize that if the maximum [b]exists[/b], there is no difference between the two. If the maximum does not exist, then you [b]have[/b] to use the supremum.
    $endgroup$
    – user247327
    Jan 3 at 18:12














1












1








1


1



$begingroup$


I have a problem with MLE's definition:



Casella Berger in Statistical Inference and Nitis Mukhopadhyay in Probability and Statistics said that MLE for a parameter $thetainTheta$ is respectively $argsup_{thetainTheta}{L(thetamid x)}$ or $argmax_{thetainTheta}{L(thetamid x)}$.



But this estimator is a supremum or a maximum? If it is a supremum why we don't call it supremum likelihood estimator? Conversely if it is a maximum, the likelihood function must be continuous and the parametric space must be compact (sufficient condition for the existence of maximum)



What is the truth or the minimal condition for the existence of MLE?






probability statistics statistical-inference estimation maximum-likelihood







share|cite|improve this question











$endgroup$




I have a problem with MLE's definition:



Casella Berger in Statistical Inference and Nitis Mukhopadhyay in Probability and Statistics said that MLE for a parameter $thetainTheta$ is respectively $argsup_{thetainTheta}{L(thetamid x)}$ or $argmax_{thetainTheta}{L(thetamid x)}$.



But this estimator is a supremum or a maximum? If it is a supremum why we don't call it supremum likelihood estimator? Conversely if it is a maximum, the likelihood function must be continuous and the parametric space must be compact (sufficient condition for the existence of maximum)



What is the truth or the minimal condition for the existence of MLE?






probability statistics statistical-inference estimation maximum-likelihood




probability statistics statistical-inference estimation maximum-likelihood






share|cite|improve this question















share|cite|improve this question













share|cite|improve this question




share|cite|improve this question








edited Jan 2 at 17:42









StubbornAtom

5,51211138




5,51211138










asked Jan 2 at 16:20









EliaElia

295




295








  • 2




    $begingroup$
    Do you understand the difference between a "supremum" and a "maximum"? Any set of real numbers, having an upper bound, as a supremum but not necessarily a maximum. A set of real numbers has a maximum if and only if the supremum is in the set in which case the maximum is the supremum.
    $endgroup$
    – user247327
    Jan 2 at 17:47










  • $begingroup$
    Of course I know the difference. My problem is that in an estimation, the estimate parameter must belong to the parametric space but if the parametric space is for example an open set, the supremum may stay in its boundary and this thing may has no sense. So I want to understand which is the correct definition of MLE; there is argsup or argmax? because the second one may not belong to the parametric space. Thanks
    $endgroup$
    – Elia
    Jan 2 at 18:54






  • 1




    $begingroup$
    Then you realize that if the maximum [b]exists[/b], there is no difference between the two. If the maximum does not exist, then you [b]have[/b] to use the supremum.
    $endgroup$
    – user247327
    Jan 3 at 18:12














  • 2




    $begingroup$
    Do you understand the difference between a "supremum" and a "maximum"? Any set of real numbers, having an upper bound, as a supremum but not necessarily a maximum. A set of real numbers has a maximum if and only if the supremum is in the set in which case the maximum is the supremum.
    $endgroup$
    – user247327
    Jan 2 at 17:47










  • $begingroup$
    Of course I know the difference. My problem is that in an estimation, the estimate parameter must belong to the parametric space but if the parametric space is for example an open set, the supremum may stay in its boundary and this thing may has no sense. So I want to understand which is the correct definition of MLE; there is argsup or argmax? because the second one may not belong to the parametric space. Thanks
    $endgroup$
    – Elia
    Jan 2 at 18:54






  • 1




    $begingroup$
    Then you realize that if the maximum [b]exists[/b], there is no difference between the two. If the maximum does not exist, then you [b]have[/b] to use the supremum.
    $endgroup$
    – user247327
    Jan 3 at 18:12








2




2




$begingroup$
Do you understand the difference between a "supremum" and a "maximum"? Any set of real numbers, having an upper bound, as a supremum but not necessarily a maximum. A set of real numbers has a maximum if and only if the supremum is in the set in which case the maximum is the supremum.
$endgroup$
– user247327
Jan 2 at 17:47




$begingroup$
Do you understand the difference between a "supremum" and a "maximum"? Any set of real numbers, having an upper bound, as a supremum but not necessarily a maximum. A set of real numbers has a maximum if and only if the supremum is in the set in which case the maximum is the supremum.
$endgroup$
– user247327
Jan 2 at 17:47












$begingroup$
Of course I know the difference. My problem is that in an estimation, the estimate parameter must belong to the parametric space but if the parametric space is for example an open set, the supremum may stay in its boundary and this thing may has no sense. So I want to understand which is the correct definition of MLE; there is argsup or argmax? because the second one may not belong to the parametric space. Thanks
$endgroup$
– Elia
Jan 2 at 18:54




$begingroup$
Of course I know the difference. My problem is that in an estimation, the estimate parameter must belong to the parametric space but if the parametric space is for example an open set, the supremum may stay in its boundary and this thing may has no sense. So I want to understand which is the correct definition of MLE; there is argsup or argmax? because the second one may not belong to the parametric space. Thanks
$endgroup$
– Elia
Jan 2 at 18:54




1




1




$begingroup$
Then you realize that if the maximum [b]exists[/b], there is no difference between the two. If the maximum does not exist, then you [b]have[/b] to use the supremum.
$endgroup$
– user247327
Jan 3 at 18:12




$begingroup$
Then you realize that if the maximum [b]exists[/b], there is no difference between the two. If the maximum does not exist, then you [b]have[/b] to use the supremum.
$endgroup$
– user247327
Jan 3 at 18:12










1 Answer
1






active

oldest

votes


















0












$begingroup$

Sufficient conditions for the MLE to exist are $L$ continuous in $theta$ over a compact support (extreme value theorem).

We can built an example where the maximum in $theta$ does not exist, but the supremum does. The main open difficulty is whether the "supremum likelihood" is meaningful...

If $U$ is uniformly distributed on $theta=(a,b)$, where $a$ and $b$ are parameters, then the likelihood (for one observation) is $$ L(theta)=frac{1}{b-a} hspace{4mm} text{ if } a<u_i<b $$ and $0$ otherwise. Then the MLE does not exist, but the SLE does (but is not unique…). See for instance the lecture notes of Songfeng Zheng (p. 5 and 6) for a discussion:
http://people.missouristate.edu/songfengzheng/Teaching/MTH541/Lecture%20notes/MLE.pdf






share|cite|improve this answer











$endgroup$













  • $begingroup$
    Thank you so much: I've searching this example and now I've understand the problem thanks to the discussion in the lecture notes.
    $endgroup$
    – Elia
    Jan 8 at 20:41











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
});


}
});














draft saved

draft discarded


















StackExchange.ready(
function () {
StackExchange.openid.initPostLogin('.new-post-login', 'https%3a%2f%2fmath.stackexchange.com%2fquestions%2f3059667%2fexistence-of-mle%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









0












$begingroup$

Sufficient conditions for the MLE to exist are $L$ continuous in $theta$ over a compact support (extreme value theorem).

We can built an example where the maximum in $theta$ does not exist, but the supremum does. The main open difficulty is whether the "supremum likelihood" is meaningful...

If $U$ is uniformly distributed on $theta=(a,b)$, where $a$ and $b$ are parameters, then the likelihood (for one observation) is $$ L(theta)=frac{1}{b-a} hspace{4mm} text{ if } a<u_i<b $$ and $0$ otherwise. Then the MLE does not exist, but the SLE does (but is not unique…). See for instance the lecture notes of Songfeng Zheng (p. 5 and 6) for a discussion:
http://people.missouristate.edu/songfengzheng/Teaching/MTH541/Lecture%20notes/MLE.pdf






share|cite|improve this answer











$endgroup$













  • $begingroup$
    Thank you so much: I've searching this example and now I've understand the problem thanks to the discussion in the lecture notes.
    $endgroup$
    – Elia
    Jan 8 at 20:41
















0












$begingroup$

Sufficient conditions for the MLE to exist are $L$ continuous in $theta$ over a compact support (extreme value theorem).

We can built an example where the maximum in $theta$ does not exist, but the supremum does. The main open difficulty is whether the "supremum likelihood" is meaningful...

If $U$ is uniformly distributed on $theta=(a,b)$, where $a$ and $b$ are parameters, then the likelihood (for one observation) is $$ L(theta)=frac{1}{b-a} hspace{4mm} text{ if } a<u_i<b $$ and $0$ otherwise. Then the MLE does not exist, but the SLE does (but is not unique…). See for instance the lecture notes of Songfeng Zheng (p. 5 and 6) for a discussion:
http://people.missouristate.edu/songfengzheng/Teaching/MTH541/Lecture%20notes/MLE.pdf






share|cite|improve this answer











$endgroup$













  • $begingroup$
    Thank you so much: I've searching this example and now I've understand the problem thanks to the discussion in the lecture notes.
    $endgroup$
    – Elia
    Jan 8 at 20:41














0












0








0





$begingroup$

Sufficient conditions for the MLE to exist are $L$ continuous in $theta$ over a compact support (extreme value theorem).

We can built an example where the maximum in $theta$ does not exist, but the supremum does. The main open difficulty is whether the "supremum likelihood" is meaningful...

If $U$ is uniformly distributed on $theta=(a,b)$, where $a$ and $b$ are parameters, then the likelihood (for one observation) is $$ L(theta)=frac{1}{b-a} hspace{4mm} text{ if } a<u_i<b $$ and $0$ otherwise. Then the MLE does not exist, but the SLE does (but is not unique…). See for instance the lecture notes of Songfeng Zheng (p. 5 and 6) for a discussion:
http://people.missouristate.edu/songfengzheng/Teaching/MTH541/Lecture%20notes/MLE.pdf






share|cite|improve this answer











$endgroup$



Sufficient conditions for the MLE to exist are $L$ continuous in $theta$ over a compact support (extreme value theorem).

We can built an example where the maximum in $theta$ does not exist, but the supremum does. The main open difficulty is whether the "supremum likelihood" is meaningful...

If $U$ is uniformly distributed on $theta=(a,b)$, where $a$ and $b$ are parameters, then the likelihood (for one observation) is $$ L(theta)=frac{1}{b-a} hspace{4mm} text{ if } a<u_i<b $$ and $0$ otherwise. Then the MLE does not exist, but the SLE does (but is not unique…). See for instance the lecture notes of Songfeng Zheng (p. 5 and 6) for a discussion:
http://people.missouristate.edu/songfengzheng/Teaching/MTH541/Lecture%20notes/MLE.pdf







share|cite|improve this answer














share|cite|improve this answer



share|cite|improve this answer








edited Jan 10 at 8:19

























answered Jan 6 at 19:41









BertrandBertrand

963




963












  • $begingroup$
    Thank you so much: I've searching this example and now I've understand the problem thanks to the discussion in the lecture notes.
    $endgroup$
    – Elia
    Jan 8 at 20:41


















  • $begingroup$
    Thank you so much: I've searching this example and now I've understand the problem thanks to the discussion in the lecture notes.
    $endgroup$
    – Elia
    Jan 8 at 20:41
















$begingroup$
Thank you so much: I've searching this example and now I've understand the problem thanks to the discussion in the lecture notes.
$endgroup$
– Elia
Jan 8 at 20:41




$begingroup$
Thank you so much: I've searching this example and now I've understand the problem thanks to the discussion in the lecture notes.
$endgroup$
– Elia
Jan 8 at 20:41


















draft saved

draft discarded




















































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.




draft saved


draft discarded














StackExchange.ready(
function () {
StackExchange.openid.initPostLogin('.new-post-login', 'https%3a%2f%2fmath.stackexchange.com%2fquestions%2f3059667%2fexistence-of-mle%23new-answer', 'question_page');
}
);

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







Popular posts from this blog

'app-layout' is not a known element: how to share Component with different Modules

Image
0 My application was working fine until I decided to go with lazy loading: So, my shared component looks like this: import { Component, Renderer2 } from '@angular/core'; export interface FormModel { captcha?: string; } @Component({ selector: 'app-layout', templateUrl: './app-layout.component.html', styleUrls: ['./app-layout.css'] }) export class FullLayoutComponent { } app-layout.component.ts I am using this component in my Another component say school-home.component.html like this: <app-layout> </app-layout> and it was working fine until I created a Module school-home.module.ts like this: import { NgModule } from '@angular/core'; import { CommonModule } from '@angular/common'; import { SchoolhomeRoutingModule ...
Read more

android studio warns about leanback feature tag usage required on manifest while using Unity exported app?

Image
.everyoneloves__top-leaderboard:empty,.everyoneloves__mid-leaderboard:empty,.everyoneloves__bot-mid-leaderboard:empty{ height:90px;width:728px;box-sizing:border-box; } 0 This is a simple project showing cubes on Unity, exported to android project and imported on Android studio. On maifest file it showed following warning: Expecting <uses-feature android:name="android.software.leanback" android:required="false" /> tag I have no information about it. And also on default emptyview android project does not use/mention this feature. I know it is required by TV's and such but why its not added by default? Why is this? And Should I concern about it? android unity3d ...
Read more

WPF add header to Image with URL pettitions [duplicate]

Image
up vote 0 down vote favorite This question already has an answer here: Image source URI requiring authorization 3 answers Add header to all* HTTP requests on a machine 1 answer first of all sorry for my english. Recently I added some security in the backend of my project, and now I need to send a validation code(apiKey) inside the headers to retrieve the images. I have a problem with ALL images in my project that I had attached in my UI. For example, before I had: <Image Height="90" Source="{Binding Video.Thumbnail}"/> Now I need to change my code to create a CUSTOM petition add the apiKey in the header:...
Read more