Optimization of the best element of the set of expressions - what kind of the optimization problem this is?












0












$begingroup$


Let the following expression describes the conditional probability of the translation of the source sequence to the target sequence (e.g. for the sentence-to-sentence translation in the neural machine translation, where xi and yj are the words in the source and target language respectively):



![P(y_{1}^{N_{T}}, theta) = prod_{i=1}^{N_{T}} P(y_{i}|y_{


This probability depends on the parameters theta of the translator function (e.g. the weights and biases of the neural network) and this probability is used as the objective function to be maximized, to find optimal parameters theta* and optimal choices yj that gives the maximum of this probability. This can be solved using gradient methods, backpropagation. That is fine.



But I am trying to consider the case when the neural machine translator generates multiple target sentences for the one source sentence. Some sentences can be complete junk, some sentences can be the parapharses of the target sentence and one sentence can be the optimal choice. So - I would like to optimize the parameters theta in such manner that the following function achieves maximum:



![max_{theta} big{P(u_{1}^{N_{T}}, theta), P(s_{1}^{N_{T}}, theta), P(t_{1}^{N_{T}}, theta), ... big}]



So - optimizer has multiple sub-objectives and the optimizer shold find one sub-objective who can achieve the maximum over all the other sub-objective and over all the theta parameters and choices of target translations u, s, t.



What kind of optimization problem this is? How to apply gradient methods to such max-function? Where to start?










share|cite|improve this question











$endgroup$

















    0












    $begingroup$


    Let the following expression describes the conditional probability of the translation of the source sequence to the target sequence (e.g. for the sentence-to-sentence translation in the neural machine translation, where xi and yj are the words in the source and target language respectively):



    ![P(y_{1}^{N_{T}}, theta) = prod_{i=1}^{N_{T}} P(y_{i}|y_{


    This probability depends on the parameters theta of the translator function (e.g. the weights and biases of the neural network) and this probability is used as the objective function to be maximized, to find optimal parameters theta* and optimal choices yj that gives the maximum of this probability. This can be solved using gradient methods, backpropagation. That is fine.



    But I am trying to consider the case when the neural machine translator generates multiple target sentences for the one source sentence. Some sentences can be complete junk, some sentences can be the parapharses of the target sentence and one sentence can be the optimal choice. So - I would like to optimize the parameters theta in such manner that the following function achieves maximum:



    ![max_{theta} big{P(u_{1}^{N_{T}}, theta), P(s_{1}^{N_{T}}, theta), P(t_{1}^{N_{T}}, theta), ... big}]



    So - optimizer has multiple sub-objectives and the optimizer shold find one sub-objective who can achieve the maximum over all the other sub-objective and over all the theta parameters and choices of target translations u, s, t.



    What kind of optimization problem this is? How to apply gradient methods to such max-function? Where to start?










    share|cite|improve this question











    $endgroup$















      0












      0








      0





      $begingroup$


      Let the following expression describes the conditional probability of the translation of the source sequence to the target sequence (e.g. for the sentence-to-sentence translation in the neural machine translation, where xi and yj are the words in the source and target language respectively):



      ![P(y_{1}^{N_{T}}, theta) = prod_{i=1}^{N_{T}} P(y_{i}|y_{


      This probability depends on the parameters theta of the translator function (e.g. the weights and biases of the neural network) and this probability is used as the objective function to be maximized, to find optimal parameters theta* and optimal choices yj that gives the maximum of this probability. This can be solved using gradient methods, backpropagation. That is fine.



      But I am trying to consider the case when the neural machine translator generates multiple target sentences for the one source sentence. Some sentences can be complete junk, some sentences can be the parapharses of the target sentence and one sentence can be the optimal choice. So - I would like to optimize the parameters theta in such manner that the following function achieves maximum:



      ![max_{theta} big{P(u_{1}^{N_{T}}, theta), P(s_{1}^{N_{T}}, theta), P(t_{1}^{N_{T}}, theta), ... big}]



      So - optimizer has multiple sub-objectives and the optimizer shold find one sub-objective who can achieve the maximum over all the other sub-objective and over all the theta parameters and choices of target translations u, s, t.



      What kind of optimization problem this is? How to apply gradient methods to such max-function? Where to start?










      share|cite|improve this question











      $endgroup$




      Let the following expression describes the conditional probability of the translation of the source sequence to the target sequence (e.g. for the sentence-to-sentence translation in the neural machine translation, where xi and yj are the words in the source and target language respectively):



      ![P(y_{1}^{N_{T}}, theta) = prod_{i=1}^{N_{T}} P(y_{i}|y_{


      This probability depends on the parameters theta of the translator function (e.g. the weights and biases of the neural network) and this probability is used as the objective function to be maximized, to find optimal parameters theta* and optimal choices yj that gives the maximum of this probability. This can be solved using gradient methods, backpropagation. That is fine.



      But I am trying to consider the case when the neural machine translator generates multiple target sentences for the one source sentence. Some sentences can be complete junk, some sentences can be the parapharses of the target sentence and one sentence can be the optimal choice. So - I would like to optimize the parameters theta in such manner that the following function achieves maximum:



      ![max_{theta} big{P(u_{1}^{N_{T}}, theta), P(s_{1}^{N_{T}}, theta), P(t_{1}^{N_{T}}, theta), ... big}]



      So - optimizer has multiple sub-objectives and the optimizer shold find one sub-objective who can achieve the maximum over all the other sub-objective and over all the theta parameters and choices of target translations u, s, t.



      What kind of optimization problem this is? How to apply gradient methods to such max-function? Where to start?







      optimization nonlinear-optimization neural-networks programming






      share|cite|improve this question















      share|cite|improve this question













      share|cite|improve this question




      share|cite|improve this question








      edited Dec 26 '18 at 14:14







      TomR

















      asked Dec 26 '18 at 13:59









      TomRTomR

      263312




      263312






















          0






          active

          oldest

          votes











          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%2f3052961%2foptimization-of-the-best-element-of-the-set-of-expressions-what-kind-of-the-op%23new-answer', 'question_page');
          }
          );

          Post as a guest















          Required, but never shown

























          0






          active

          oldest

          votes








          0






          active

          oldest

          votes









          active

          oldest

          votes






          active

          oldest

          votes
















          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%2f3052961%2foptimization-of-the-best-element-of-the-set-of-expressions-what-kind-of-the-op%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

          Ellipse (mathématiques)

          Quarter-circle Tiles

          Mont Emei