Please correct my thinking about Ridge Regression












0














If ridge regression biases ALL beta coefficients of a regression model towards zero, wouldn't the model massively mispredict the y-variable?



I know my logic must be wrong here, but I'd appreciate if someone could point out the flaw. I am just starting to get into this topic, so would appreciate as simple an answer as possible.



For example: Say you have the coefficients of a linear regression output (and for simplicity, all coefficients are positive, and say the x-variables can only take on positive values). Then, if you biased each beta coefficient towards zero, wouldn't the model consistently underpredict the y-variable?



Would appreciate any intuition on how ridge regression shrinks all the coefficients but still maintains predictive ability. Thank you!










share|cite|improve this question



























    0














    If ridge regression biases ALL beta coefficients of a regression model towards zero, wouldn't the model massively mispredict the y-variable?



    I know my logic must be wrong here, but I'd appreciate if someone could point out the flaw. I am just starting to get into this topic, so would appreciate as simple an answer as possible.



    For example: Say you have the coefficients of a linear regression output (and for simplicity, all coefficients are positive, and say the x-variables can only take on positive values). Then, if you biased each beta coefficient towards zero, wouldn't the model consistently underpredict the y-variable?



    Would appreciate any intuition on how ridge regression shrinks all the coefficients but still maintains predictive ability. Thank you!










    share|cite|improve this question

























      0












      0








      0







      If ridge regression biases ALL beta coefficients of a regression model towards zero, wouldn't the model massively mispredict the y-variable?



      I know my logic must be wrong here, but I'd appreciate if someone could point out the flaw. I am just starting to get into this topic, so would appreciate as simple an answer as possible.



      For example: Say you have the coefficients of a linear regression output (and for simplicity, all coefficients are positive, and say the x-variables can only take on positive values). Then, if you biased each beta coefficient towards zero, wouldn't the model consistently underpredict the y-variable?



      Would appreciate any intuition on how ridge regression shrinks all the coefficients but still maintains predictive ability. Thank you!










      share|cite|improve this question













      If ridge regression biases ALL beta coefficients of a regression model towards zero, wouldn't the model massively mispredict the y-variable?



      I know my logic must be wrong here, but I'd appreciate if someone could point out the flaw. I am just starting to get into this topic, so would appreciate as simple an answer as possible.



      For example: Say you have the coefficients of a linear regression output (and for simplicity, all coefficients are positive, and say the x-variables can only take on positive values). Then, if you biased each beta coefficient towards zero, wouldn't the model consistently underpredict the y-variable?



      Would appreciate any intuition on how ridge regression shrinks all the coefficients but still maintains predictive ability. Thank you!







      statistics regression mathematical-modeling linear-regression regression-analysis






      share|cite|improve this question













      share|cite|improve this question











      share|cite|improve this question




      share|cite|improve this question










      asked Dec 2 at 17:18









      bob

      1089




      1089






















          1 Answer
          1






          active

          oldest

          votes


















          0














          The objective function in ridge regression has two terms. You are correct that the penalty term $lambda |beta|_2^2$ encourages the coefficients to be small. However, you forget that the main term $|y - X beta|^2$ encourages the choice of coefficients to fit $y$ well. So there is a tension between these two competing goals of fitting $y$ well and keeping the coefficients small.



          Regarding "constantly underpredict," you are correct that ridge regression is biased. But this is not the only criteria for evaluating the performance of an estimator (bias-variance tradeoff).






          share|cite|improve this answer





















            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%2f3022920%2fplease-correct-my-thinking-about-ridge-regression%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














            The objective function in ridge regression has two terms. You are correct that the penalty term $lambda |beta|_2^2$ encourages the coefficients to be small. However, you forget that the main term $|y - X beta|^2$ encourages the choice of coefficients to fit $y$ well. So there is a tension between these two competing goals of fitting $y$ well and keeping the coefficients small.



            Regarding "constantly underpredict," you are correct that ridge regression is biased. But this is not the only criteria for evaluating the performance of an estimator (bias-variance tradeoff).






            share|cite|improve this answer


























              0














              The objective function in ridge regression has two terms. You are correct that the penalty term $lambda |beta|_2^2$ encourages the coefficients to be small. However, you forget that the main term $|y - X beta|^2$ encourages the choice of coefficients to fit $y$ well. So there is a tension between these two competing goals of fitting $y$ well and keeping the coefficients small.



              Regarding "constantly underpredict," you are correct that ridge regression is biased. But this is not the only criteria for evaluating the performance of an estimator (bias-variance tradeoff).






              share|cite|improve this answer
























                0












                0








                0






                The objective function in ridge regression has two terms. You are correct that the penalty term $lambda |beta|_2^2$ encourages the coefficients to be small. However, you forget that the main term $|y - X beta|^2$ encourages the choice of coefficients to fit $y$ well. So there is a tension between these two competing goals of fitting $y$ well and keeping the coefficients small.



                Regarding "constantly underpredict," you are correct that ridge regression is biased. But this is not the only criteria for evaluating the performance of an estimator (bias-variance tradeoff).






                share|cite|improve this answer












                The objective function in ridge regression has two terms. You are correct that the penalty term $lambda |beta|_2^2$ encourages the coefficients to be small. However, you forget that the main term $|y - X beta|^2$ encourages the choice of coefficients to fit $y$ well. So there is a tension between these two competing goals of fitting $y$ well and keeping the coefficients small.



                Regarding "constantly underpredict," you are correct that ridge regression is biased. But this is not the only criteria for evaluating the performance of an estimator (bias-variance tradeoff).







                share|cite|improve this answer












                share|cite|improve this answer



                share|cite|improve this answer










                answered Dec 2 at 17:43









                angryavian

                38.9k23180




                38.9k23180






























                    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.





                    Some of your past answers have not been well-received, and you're in danger of being blocked from answering.


                    Please pay close attention to the following guidance:


                    • 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.


                    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%2f3022920%2fplease-correct-my-thinking-about-ridge-regression%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

                    Berounka

                    Sphinx de Gizeh

                    Different font size/position of beamer's navigation symbols template's content depending on regular/plain...