How to show $text{Tr}(Mlog N)=sum_{i,j}^nlambda_ilog(tilde{lambda_j})(u_i^{top}tilde{u}_j)^2$?












3












$begingroup$


The above question is the equation $(2.4)$ of the following paper:



MATRIX EXPONENTIATED GRADIENT UPDATES.



Let $M$ and $N$ be two $n times n$ positive definite matrices where $M=ULambda U^{top}$, $N=tilde{U}tilde{Lambda} tilde{U}^{top}
$
and $(lambda_i,v_i)$ are eigenpairs of $M$, likewise for $N$.



How to show the following
$$text{Tr}(Mlog N)=sum_{i,j}lambda_ilog(tilde{lambda_j})(u_i^{top}tilde{u}_j)^2$$



First I do not know what $i,j$ mean in summation, and how we have it using two summations. Second how to get that.



My try:



begin{align}
text{Tr}(Mlog N) &=
text{Tr}(ULambda U^{top}
tilde{U}log(tilde{Lambda})tilde{U}^{top}) \
& = text{Tr}(Lambda U^{top}
tilde{U}log(tilde{Lambda})tilde{U}^{top}U)
end{align}



How can I proceed using matrix calculus to get the result not by expanding? what is the hidden trick?










share|cite|improve this question











$endgroup$

















    3












    $begingroup$


    The above question is the equation $(2.4)$ of the following paper:



    MATRIX EXPONENTIATED GRADIENT UPDATES.



    Let $M$ and $N$ be two $n times n$ positive definite matrices where $M=ULambda U^{top}$, $N=tilde{U}tilde{Lambda} tilde{U}^{top}
    $
    and $(lambda_i,v_i)$ are eigenpairs of $M$, likewise for $N$.



    How to show the following
    $$text{Tr}(Mlog N)=sum_{i,j}lambda_ilog(tilde{lambda_j})(u_i^{top}tilde{u}_j)^2$$



    First I do not know what $i,j$ mean in summation, and how we have it using two summations. Second how to get that.



    My try:



    begin{align}
    text{Tr}(Mlog N) &=
    text{Tr}(ULambda U^{top}
    tilde{U}log(tilde{Lambda})tilde{U}^{top}) \
    & = text{Tr}(Lambda U^{top}
    tilde{U}log(tilde{Lambda})tilde{U}^{top}U)
    end{align}



    How can I proceed using matrix calculus to get the result not by expanding? what is the hidden trick?










    share|cite|improve this question











    $endgroup$















      3












      3








      3





      $begingroup$


      The above question is the equation $(2.4)$ of the following paper:



      MATRIX EXPONENTIATED GRADIENT UPDATES.



      Let $M$ and $N$ be two $n times n$ positive definite matrices where $M=ULambda U^{top}$, $N=tilde{U}tilde{Lambda} tilde{U}^{top}
      $
      and $(lambda_i,v_i)$ are eigenpairs of $M$, likewise for $N$.



      How to show the following
      $$text{Tr}(Mlog N)=sum_{i,j}lambda_ilog(tilde{lambda_j})(u_i^{top}tilde{u}_j)^2$$



      First I do not know what $i,j$ mean in summation, and how we have it using two summations. Second how to get that.



      My try:



      begin{align}
      text{Tr}(Mlog N) &=
      text{Tr}(ULambda U^{top}
      tilde{U}log(tilde{Lambda})tilde{U}^{top}) \
      & = text{Tr}(Lambda U^{top}
      tilde{U}log(tilde{Lambda})tilde{U}^{top}U)
      end{align}



      How can I proceed using matrix calculus to get the result not by expanding? what is the hidden trick?










      share|cite|improve this question











      $endgroup$




      The above question is the equation $(2.4)$ of the following paper:



      MATRIX EXPONENTIATED GRADIENT UPDATES.



      Let $M$ and $N$ be two $n times n$ positive definite matrices where $M=ULambda U^{top}$, $N=tilde{U}tilde{Lambda} tilde{U}^{top}
      $
      and $(lambda_i,v_i)$ are eigenpairs of $M$, likewise for $N$.



      How to show the following
      $$text{Tr}(Mlog N)=sum_{i,j}lambda_ilog(tilde{lambda_j})(u_i^{top}tilde{u}_j)^2$$



      First I do not know what $i,j$ mean in summation, and how we have it using two summations. Second how to get that.



      My try:



      begin{align}
      text{Tr}(Mlog N) &=
      text{Tr}(ULambda U^{top}
      tilde{U}log(tilde{Lambda})tilde{U}^{top}) \
      & = text{Tr}(Lambda U^{top}
      tilde{U}log(tilde{Lambda})tilde{U}^{top}U)
      end{align}



      How can I proceed using matrix calculus to get the result not by expanding? what is the hidden trick?







      linear-algebra matrices positive-definite symmetric-matrices






      share|cite|improve this question















      share|cite|improve this question













      share|cite|improve this question




      share|cite|improve this question








      edited Jan 20 at 3:36









      loup blanc

      23.5k21851




      23.5k21851










      asked Jan 18 at 20:17









      SaeedSaeed

      1,073310




      1,073310






















          2 Answers
          2






          active

          oldest

          votes


















          3












          $begingroup$

          To easy the notations, let us take
          $$
          kappa_j=logtilde{lambda}_j
          $$

          and
          $$
          v_j=tilde{u}_j.
          $$

          With these notations, it suffices to show
          $$
          text{tr}left(Mlog Nright)=sum_{i,j=1}^nlambda_ikappa_jleft(u_i^{top}v_jright)^2.
          $$



          Now, let us begin our proof. Using
          begin{align}
          M&=sum_{j=1}^nlambda_ju_ju_j^{top},\
          N&=sum_{j=1}^nkappa_jv_jv_j^{top},
          end{align}

          we have
          begin{align}
          text{tr}left(Mlog Nright)&=text{tr}left(sum_{i=1}^nlambda_iu_iu_i^{top}sum_{j=1}^nkappa_jv_jv_j^{top}right)\
          &=text{tr}left(sum_{i,j=1}^nlambda_ikappa_ju_iu_i^{top}v_jv_j^{top}right)\
          &=sum_{i,j=1}^nlambda_ikappa_jtext{tr}left(u_iu_i^{top}v_jv_j^{top}right)\
          &=sum_{i,j=1}^nlambda_ikappa_jtext{tr}left(v_j^{top}u_iu_i^{top}v_jright)\
          &=sum_{i,j=1}^nlambda_ikappa_jleft(v_j^{top}u_iright)left(u_i^{top}v_jright)\
          &=sum_{i,j=1}^nlambda_ikappa_jleft(u_i^{top}v_jright)^2.
          end{align}

          This completes the proof.



          I would like to thank @loupblanc for kind advices.






          share|cite|improve this answer











          $endgroup$













          • $begingroup$
            Your proof is correct. 2 remarks: i) write $log(N)=cdots$ and ii) using Frobenius is useless because $u_i^Tv_j=v_j^Tu_i$.
            $endgroup$
            – loup blanc
            Jan 20 at 3:35










          • $begingroup$
            @hypernova: Does the summation have $n^2$ terms?
            $endgroup$
            – Saeed
            Jan 20 at 4:27










          • $begingroup$
            @loupblanc: Thank you for your advices. I modified my answer as per your suggestions, credited to you.
            $endgroup$
            – hypernova
            Jan 20 at 5:11










          • $begingroup$
            @Saeed: Yes, the summation contains $n^2$ terms, and each of the indices $i$ and $j$ runs from $1$ to $n$. In other words, $sum_{i,j=1}^n=sum_{i=1}^nsum_{j=1}^n$.
            $endgroup$
            – hypernova
            Jan 20 at 5:11





















          2












          $begingroup$

          The first version of the formula with one index is false. Take $M=begin{pmatrix}50&37\37&61end{pmatrix},N=diag(2,1)$.






          share|cite|improve this answer











          $endgroup$













          • $begingroup$
            It is on page 999 of the following paper jmlr.org/papers/volume6/tsuda05a/tsuda05a.pdf. Maybe I am wrong. Because I am confused about $i$ and $j$.
            $endgroup$
            – Saeed
            Jan 20 at 1:29










          • $begingroup$
            Yes, the correct formula is a sum with $2$ indices.
            $endgroup$
            – loup blanc
            Jan 20 at 3:38











          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%2f3078707%2fhow-to-show-texttrm-log-n-sum-i-jn-lambda-i-log-tilde-lambda-ju-i%23new-answer', 'question_page');
          }
          );

          Post as a guest















          Required, but never shown

























          2 Answers
          2






          active

          oldest

          votes








          2 Answers
          2






          active

          oldest

          votes









          active

          oldest

          votes






          active

          oldest

          votes









          3












          $begingroup$

          To easy the notations, let us take
          $$
          kappa_j=logtilde{lambda}_j
          $$

          and
          $$
          v_j=tilde{u}_j.
          $$

          With these notations, it suffices to show
          $$
          text{tr}left(Mlog Nright)=sum_{i,j=1}^nlambda_ikappa_jleft(u_i^{top}v_jright)^2.
          $$



          Now, let us begin our proof. Using
          begin{align}
          M&=sum_{j=1}^nlambda_ju_ju_j^{top},\
          N&=sum_{j=1}^nkappa_jv_jv_j^{top},
          end{align}

          we have
          begin{align}
          text{tr}left(Mlog Nright)&=text{tr}left(sum_{i=1}^nlambda_iu_iu_i^{top}sum_{j=1}^nkappa_jv_jv_j^{top}right)\
          &=text{tr}left(sum_{i,j=1}^nlambda_ikappa_ju_iu_i^{top}v_jv_j^{top}right)\
          &=sum_{i,j=1}^nlambda_ikappa_jtext{tr}left(u_iu_i^{top}v_jv_j^{top}right)\
          &=sum_{i,j=1}^nlambda_ikappa_jtext{tr}left(v_j^{top}u_iu_i^{top}v_jright)\
          &=sum_{i,j=1}^nlambda_ikappa_jleft(v_j^{top}u_iright)left(u_i^{top}v_jright)\
          &=sum_{i,j=1}^nlambda_ikappa_jleft(u_i^{top}v_jright)^2.
          end{align}

          This completes the proof.



          I would like to thank @loupblanc for kind advices.






          share|cite|improve this answer











          $endgroup$













          • $begingroup$
            Your proof is correct. 2 remarks: i) write $log(N)=cdots$ and ii) using Frobenius is useless because $u_i^Tv_j=v_j^Tu_i$.
            $endgroup$
            – loup blanc
            Jan 20 at 3:35










          • $begingroup$
            @hypernova: Does the summation have $n^2$ terms?
            $endgroup$
            – Saeed
            Jan 20 at 4:27










          • $begingroup$
            @loupblanc: Thank you for your advices. I modified my answer as per your suggestions, credited to you.
            $endgroup$
            – hypernova
            Jan 20 at 5:11










          • $begingroup$
            @Saeed: Yes, the summation contains $n^2$ terms, and each of the indices $i$ and $j$ runs from $1$ to $n$. In other words, $sum_{i,j=1}^n=sum_{i=1}^nsum_{j=1}^n$.
            $endgroup$
            – hypernova
            Jan 20 at 5:11


















          3












          $begingroup$

          To easy the notations, let us take
          $$
          kappa_j=logtilde{lambda}_j
          $$

          and
          $$
          v_j=tilde{u}_j.
          $$

          With these notations, it suffices to show
          $$
          text{tr}left(Mlog Nright)=sum_{i,j=1}^nlambda_ikappa_jleft(u_i^{top}v_jright)^2.
          $$



          Now, let us begin our proof. Using
          begin{align}
          M&=sum_{j=1}^nlambda_ju_ju_j^{top},\
          N&=sum_{j=1}^nkappa_jv_jv_j^{top},
          end{align}

          we have
          begin{align}
          text{tr}left(Mlog Nright)&=text{tr}left(sum_{i=1}^nlambda_iu_iu_i^{top}sum_{j=1}^nkappa_jv_jv_j^{top}right)\
          &=text{tr}left(sum_{i,j=1}^nlambda_ikappa_ju_iu_i^{top}v_jv_j^{top}right)\
          &=sum_{i,j=1}^nlambda_ikappa_jtext{tr}left(u_iu_i^{top}v_jv_j^{top}right)\
          &=sum_{i,j=1}^nlambda_ikappa_jtext{tr}left(v_j^{top}u_iu_i^{top}v_jright)\
          &=sum_{i,j=1}^nlambda_ikappa_jleft(v_j^{top}u_iright)left(u_i^{top}v_jright)\
          &=sum_{i,j=1}^nlambda_ikappa_jleft(u_i^{top}v_jright)^2.
          end{align}

          This completes the proof.



          I would like to thank @loupblanc for kind advices.






          share|cite|improve this answer











          $endgroup$













          • $begingroup$
            Your proof is correct. 2 remarks: i) write $log(N)=cdots$ and ii) using Frobenius is useless because $u_i^Tv_j=v_j^Tu_i$.
            $endgroup$
            – loup blanc
            Jan 20 at 3:35










          • $begingroup$
            @hypernova: Does the summation have $n^2$ terms?
            $endgroup$
            – Saeed
            Jan 20 at 4:27










          • $begingroup$
            @loupblanc: Thank you for your advices. I modified my answer as per your suggestions, credited to you.
            $endgroup$
            – hypernova
            Jan 20 at 5:11










          • $begingroup$
            @Saeed: Yes, the summation contains $n^2$ terms, and each of the indices $i$ and $j$ runs from $1$ to $n$. In other words, $sum_{i,j=1}^n=sum_{i=1}^nsum_{j=1}^n$.
            $endgroup$
            – hypernova
            Jan 20 at 5:11
















          3












          3








          3





          $begingroup$

          To easy the notations, let us take
          $$
          kappa_j=logtilde{lambda}_j
          $$

          and
          $$
          v_j=tilde{u}_j.
          $$

          With these notations, it suffices to show
          $$
          text{tr}left(Mlog Nright)=sum_{i,j=1}^nlambda_ikappa_jleft(u_i^{top}v_jright)^2.
          $$



          Now, let us begin our proof. Using
          begin{align}
          M&=sum_{j=1}^nlambda_ju_ju_j^{top},\
          N&=sum_{j=1}^nkappa_jv_jv_j^{top},
          end{align}

          we have
          begin{align}
          text{tr}left(Mlog Nright)&=text{tr}left(sum_{i=1}^nlambda_iu_iu_i^{top}sum_{j=1}^nkappa_jv_jv_j^{top}right)\
          &=text{tr}left(sum_{i,j=1}^nlambda_ikappa_ju_iu_i^{top}v_jv_j^{top}right)\
          &=sum_{i,j=1}^nlambda_ikappa_jtext{tr}left(u_iu_i^{top}v_jv_j^{top}right)\
          &=sum_{i,j=1}^nlambda_ikappa_jtext{tr}left(v_j^{top}u_iu_i^{top}v_jright)\
          &=sum_{i,j=1}^nlambda_ikappa_jleft(v_j^{top}u_iright)left(u_i^{top}v_jright)\
          &=sum_{i,j=1}^nlambda_ikappa_jleft(u_i^{top}v_jright)^2.
          end{align}

          This completes the proof.



          I would like to thank @loupblanc for kind advices.






          share|cite|improve this answer











          $endgroup$



          To easy the notations, let us take
          $$
          kappa_j=logtilde{lambda}_j
          $$

          and
          $$
          v_j=tilde{u}_j.
          $$

          With these notations, it suffices to show
          $$
          text{tr}left(Mlog Nright)=sum_{i,j=1}^nlambda_ikappa_jleft(u_i^{top}v_jright)^2.
          $$



          Now, let us begin our proof. Using
          begin{align}
          M&=sum_{j=1}^nlambda_ju_ju_j^{top},\
          N&=sum_{j=1}^nkappa_jv_jv_j^{top},
          end{align}

          we have
          begin{align}
          text{tr}left(Mlog Nright)&=text{tr}left(sum_{i=1}^nlambda_iu_iu_i^{top}sum_{j=1}^nkappa_jv_jv_j^{top}right)\
          &=text{tr}left(sum_{i,j=1}^nlambda_ikappa_ju_iu_i^{top}v_jv_j^{top}right)\
          &=sum_{i,j=1}^nlambda_ikappa_jtext{tr}left(u_iu_i^{top}v_jv_j^{top}right)\
          &=sum_{i,j=1}^nlambda_ikappa_jtext{tr}left(v_j^{top}u_iu_i^{top}v_jright)\
          &=sum_{i,j=1}^nlambda_ikappa_jleft(v_j^{top}u_iright)left(u_i^{top}v_jright)\
          &=sum_{i,j=1}^nlambda_ikappa_jleft(u_i^{top}v_jright)^2.
          end{align}

          This completes the proof.



          I would like to thank @loupblanc for kind advices.







          share|cite|improve this answer














          share|cite|improve this answer



          share|cite|improve this answer








          edited Jan 20 at 5:10

























          answered Jan 20 at 1:42









          hypernovahypernova

          4,834414




          4,834414












          • $begingroup$
            Your proof is correct. 2 remarks: i) write $log(N)=cdots$ and ii) using Frobenius is useless because $u_i^Tv_j=v_j^Tu_i$.
            $endgroup$
            – loup blanc
            Jan 20 at 3:35










          • $begingroup$
            @hypernova: Does the summation have $n^2$ terms?
            $endgroup$
            – Saeed
            Jan 20 at 4:27










          • $begingroup$
            @loupblanc: Thank you for your advices. I modified my answer as per your suggestions, credited to you.
            $endgroup$
            – hypernova
            Jan 20 at 5:11










          • $begingroup$
            @Saeed: Yes, the summation contains $n^2$ terms, and each of the indices $i$ and $j$ runs from $1$ to $n$. In other words, $sum_{i,j=1}^n=sum_{i=1}^nsum_{j=1}^n$.
            $endgroup$
            – hypernova
            Jan 20 at 5:11




















          • $begingroup$
            Your proof is correct. 2 remarks: i) write $log(N)=cdots$ and ii) using Frobenius is useless because $u_i^Tv_j=v_j^Tu_i$.
            $endgroup$
            – loup blanc
            Jan 20 at 3:35










          • $begingroup$
            @hypernova: Does the summation have $n^2$ terms?
            $endgroup$
            – Saeed
            Jan 20 at 4:27










          • $begingroup$
            @loupblanc: Thank you for your advices. I modified my answer as per your suggestions, credited to you.
            $endgroup$
            – hypernova
            Jan 20 at 5:11










          • $begingroup$
            @Saeed: Yes, the summation contains $n^2$ terms, and each of the indices $i$ and $j$ runs from $1$ to $n$. In other words, $sum_{i,j=1}^n=sum_{i=1}^nsum_{j=1}^n$.
            $endgroup$
            – hypernova
            Jan 20 at 5:11


















          $begingroup$
          Your proof is correct. 2 remarks: i) write $log(N)=cdots$ and ii) using Frobenius is useless because $u_i^Tv_j=v_j^Tu_i$.
          $endgroup$
          – loup blanc
          Jan 20 at 3:35




          $begingroup$
          Your proof is correct. 2 remarks: i) write $log(N)=cdots$ and ii) using Frobenius is useless because $u_i^Tv_j=v_j^Tu_i$.
          $endgroup$
          – loup blanc
          Jan 20 at 3:35












          $begingroup$
          @hypernova: Does the summation have $n^2$ terms?
          $endgroup$
          – Saeed
          Jan 20 at 4:27




          $begingroup$
          @hypernova: Does the summation have $n^2$ terms?
          $endgroup$
          – Saeed
          Jan 20 at 4:27












          $begingroup$
          @loupblanc: Thank you for your advices. I modified my answer as per your suggestions, credited to you.
          $endgroup$
          – hypernova
          Jan 20 at 5:11




          $begingroup$
          @loupblanc: Thank you for your advices. I modified my answer as per your suggestions, credited to you.
          $endgroup$
          – hypernova
          Jan 20 at 5:11












          $begingroup$
          @Saeed: Yes, the summation contains $n^2$ terms, and each of the indices $i$ and $j$ runs from $1$ to $n$. In other words, $sum_{i,j=1}^n=sum_{i=1}^nsum_{j=1}^n$.
          $endgroup$
          – hypernova
          Jan 20 at 5:11






          $begingroup$
          @Saeed: Yes, the summation contains $n^2$ terms, and each of the indices $i$ and $j$ runs from $1$ to $n$. In other words, $sum_{i,j=1}^n=sum_{i=1}^nsum_{j=1}^n$.
          $endgroup$
          – hypernova
          Jan 20 at 5:11













          2












          $begingroup$

          The first version of the formula with one index is false. Take $M=begin{pmatrix}50&37\37&61end{pmatrix},N=diag(2,1)$.






          share|cite|improve this answer











          $endgroup$













          • $begingroup$
            It is on page 999 of the following paper jmlr.org/papers/volume6/tsuda05a/tsuda05a.pdf. Maybe I am wrong. Because I am confused about $i$ and $j$.
            $endgroup$
            – Saeed
            Jan 20 at 1:29










          • $begingroup$
            Yes, the correct formula is a sum with $2$ indices.
            $endgroup$
            – loup blanc
            Jan 20 at 3:38
















          2












          $begingroup$

          The first version of the formula with one index is false. Take $M=begin{pmatrix}50&37\37&61end{pmatrix},N=diag(2,1)$.






          share|cite|improve this answer











          $endgroup$













          • $begingroup$
            It is on page 999 of the following paper jmlr.org/papers/volume6/tsuda05a/tsuda05a.pdf. Maybe I am wrong. Because I am confused about $i$ and $j$.
            $endgroup$
            – Saeed
            Jan 20 at 1:29










          • $begingroup$
            Yes, the correct formula is a sum with $2$ indices.
            $endgroup$
            – loup blanc
            Jan 20 at 3:38














          2












          2








          2





          $begingroup$

          The first version of the formula with one index is false. Take $M=begin{pmatrix}50&37\37&61end{pmatrix},N=diag(2,1)$.






          share|cite|improve this answer











          $endgroup$



          The first version of the formula with one index is false. Take $M=begin{pmatrix}50&37\37&61end{pmatrix},N=diag(2,1)$.







          share|cite|improve this answer














          share|cite|improve this answer



          share|cite|improve this answer








          edited Jan 20 at 3:41

























          answered Jan 19 at 10:04









          loup blancloup blanc

          23.5k21851




          23.5k21851












          • $begingroup$
            It is on page 999 of the following paper jmlr.org/papers/volume6/tsuda05a/tsuda05a.pdf. Maybe I am wrong. Because I am confused about $i$ and $j$.
            $endgroup$
            – Saeed
            Jan 20 at 1:29










          • $begingroup$
            Yes, the correct formula is a sum with $2$ indices.
            $endgroup$
            – loup blanc
            Jan 20 at 3:38


















          • $begingroup$
            It is on page 999 of the following paper jmlr.org/papers/volume6/tsuda05a/tsuda05a.pdf. Maybe I am wrong. Because I am confused about $i$ and $j$.
            $endgroup$
            – Saeed
            Jan 20 at 1:29










          • $begingroup$
            Yes, the correct formula is a sum with $2$ indices.
            $endgroup$
            – loup blanc
            Jan 20 at 3:38
















          $begingroup$
          It is on page 999 of the following paper jmlr.org/papers/volume6/tsuda05a/tsuda05a.pdf. Maybe I am wrong. Because I am confused about $i$ and $j$.
          $endgroup$
          – Saeed
          Jan 20 at 1:29




          $begingroup$
          It is on page 999 of the following paper jmlr.org/papers/volume6/tsuda05a/tsuda05a.pdf. Maybe I am wrong. Because I am confused about $i$ and $j$.
          $endgroup$
          – Saeed
          Jan 20 at 1:29












          $begingroup$
          Yes, the correct formula is a sum with $2$ indices.
          $endgroup$
          – loup blanc
          Jan 20 at 3:38




          $begingroup$
          Yes, the correct formula is a sum with $2$ indices.
          $endgroup$
          – loup blanc
          Jan 20 at 3:38


















          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%2f3078707%2fhow-to-show-texttrm-log-n-sum-i-jn-lambda-i-log-tilde-lambda-ju-i%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

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

          SQL update select statement

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