http://git-wip-us.apache.org/repos/asf/incubator-hivemall-site/blob/d9012d92/userguide/spark/regression/e2006_df.html ---------------------------------------------------------------------- diff --git a/userguide/spark/regression/e2006_df.html b/userguide/spark/regression/e2006_df.html index b1e1818..db96ccc 100644 --- a/userguide/spark/regression/e2006_df.html +++ b/userguide/spark/regression/e2006_df.html @@ -4,7 +4,7 @@ <head> <meta charset="UTF-8"> <meta content="text/html; charset=utf-8" http-equiv="Content-Type"> - <title>E2006-tfidf regression tutorial for DataFrame · Hivemall User Manual</title> + <title>E2006-tfidf Regression Tutorial for DataFrame · Hivemall User Manual</title> <meta http-equiv="X-UA-Compatible" content="IE=edge" /> <meta name="description" content=""> <meta name="generator" content="GitBook 3.2.3"> @@ -972,7 +972,7 @@ <b>6.2.1.</b> - Data preparation + Data Preparation </a> @@ -980,13 +980,28 @@ </li> - <li class="chapter " data-level="6.2.2" data-path="../../binaryclass/a9a_lr.html"> + <li class="chapter " data-level="6.2.2" data-path="../../binaryclass/a9a_generic.html"> - <a href="../../binaryclass/a9a_lr.html"> + <a href="../../binaryclass/a9a_generic.html"> <b>6.2.2.</b> + General Binary Classifier + + </a> + + + + </li> + + <li class="chapter " data-level="6.2.3" data-path="../../binaryclass/a9a_lr.html"> + + <a href="../../binaryclass/a9a_lr.html"> + + + <b>6.2.3.</b> + Logistic Regression </a> @@ -995,14 +1010,14 @@ </li> - <li class="chapter " data-level="6.2.3" data-path="../../binaryclass/a9a_minibatch.html"> + <li class="chapter " data-level="6.2.4" data-path="../../binaryclass/a9a_minibatch.html"> <a href="../../binaryclass/a9a_minibatch.html"> - <b>6.2.3.</b> + <b>6.2.4.</b> - Mini-batch gradient descent + Mini-batch Gradient Descent </a> @@ -1038,7 +1053,7 @@ <b>6.3.1.</b> - Data preparation + Data Preparation </a> @@ -1076,13 +1091,28 @@ </li> - <li class="chapter " data-level="6.3.4" data-path="../../binaryclass/news20_adagrad.html"> + <li class="chapter " data-level="6.3.4" data-path="../../binaryclass/news20_generic.html"> - <a href="../../binaryclass/news20_adagrad.html"> + <a href="../../binaryclass/news20_generic.html"> <b>6.3.4.</b> + General Binary Classifier + + </a> + + + + </li> + + <li class="chapter " data-level="6.3.5" data-path="../../binaryclass/news20_adagrad.html"> + + <a href="../../binaryclass/news20_adagrad.html"> + + + <b>6.3.5.</b> + AdaGradRDA, AdaGrad, AdaDelta </a> @@ -1091,12 +1121,12 @@ </li> - <li class="chapter " data-level="6.3.5" data-path="../../binaryclass/news20_rf.html"> + <li class="chapter " data-level="6.3.6" data-path="../../binaryclass/news20_rf.html"> <a href="../../binaryclass/news20_rf.html"> - <b>6.3.5.</b> + <b>6.3.6.</b> Random Forest @@ -1134,7 +1164,7 @@ <b>6.4.1.</b> - Data preparation + Data Preparation </a> @@ -1185,7 +1215,7 @@ <b>6.5.1.</b> - Data preparation + Data Preparation </a> @@ -1236,7 +1266,7 @@ <b>6.6.1.</b> - Data pareparation + Data Pareparation </a> @@ -1302,7 +1332,7 @@ <b>6.8.1.</b> - Data preparation + Data Preparation </a> @@ -1360,7 +1390,7 @@ <b>7.1.1.</b> - Data preparation + Data Preparation </a> @@ -1375,7 +1405,7 @@ <b>7.1.2.</b> - Data preparation for one-vs-the-rest classifiers + Data Preparation for one-vs-the-rest classifiers </a> @@ -1435,7 +1465,7 @@ <b>7.1.6.</b> - one-vs-the-rest classifier + one-vs-the-rest Classifier </a> @@ -1559,7 +1589,7 @@ <b>8.2.1.</b> - Data preparation + Data Preparation </a> @@ -1567,13 +1597,28 @@ </li> - <li class="chapter " data-level="8.2.2" data-path="../../regression/e2006_arow.html"> + <li class="chapter " data-level="8.2.2" data-path="../../regression/e2006_generic.html"> - <a href="../../regression/e2006_arow.html"> + <a href="../../regression/e2006_generic.html"> <b>8.2.2.</b> + General Regessor + + </a> + + + + </li> + + <li class="chapter " data-level="8.2.3" data-path="../../regression/e2006_arow.html"> + + <a href="../../regression/e2006_arow.html"> + + + <b>8.2.3.</b> + Passive Aggressive, AROW </a> @@ -1610,7 +1655,7 @@ <b>8.3.1.</b> - Data preparation + Data Preparation </a> @@ -1698,7 +1743,7 @@ <b>9.1.1.</b> - Item-based collaborative filtering + Item-based Collaborative Filtering </a> @@ -1734,7 +1779,7 @@ <b>9.2.1.</b> - Data preparation + Data Preparation </a> @@ -1749,7 +1794,7 @@ <b>9.2.2.</b> - LSH/MinHash and Jaccard similarity + LSH/MinHash and Jaccard Similarity </a> @@ -1764,7 +1809,7 @@ <b>9.2.3.</b> - LSH/MinHash and brute-force search + LSH/MinHash and Brute-force Search </a> @@ -1815,7 +1860,7 @@ <b>9.3.1.</b> - Data preparation + Data Preparation </a> @@ -1830,7 +1875,7 @@ <b>9.3.2.</b> - Item-based collaborative filtering + Item-based Collaborative Filtering </a> @@ -1875,7 +1920,7 @@ <b>9.3.5.</b> - SLIM for fast top-k recommendation + SLIM for fast top-k Recommendation </a> @@ -1890,7 +1935,7 @@ <b>9.3.6.</b> - 10-fold cross validation (Matrix Factorization) + 10-fold Cross Validation (Matrix Factorization) </a> @@ -2080,7 +2125,7 @@ <b>13.2.1.</b> - a9a tutorial for DataFrame + a9a Tutorial for DataFrame </a> @@ -2095,7 +2140,7 @@ <b>13.2.2.</b> - a9a tutorial for SQL + a9a Tutorial for SQL </a> @@ -2131,7 +2176,7 @@ <b>13.3.1.</b> - E2006-tfidf regression tutorial for DataFrame + E2006-tfidf Regression Tutorial for DataFrame </a> @@ -2146,7 +2191,7 @@ <b>13.3.2.</b> - E2006-tfidf regression tutorial for SQL + E2006-tfidf Regression Tutorial for SQL </a> @@ -2166,7 +2211,7 @@ <b>13.4.</b> - Generic features + Generic Features </a> @@ -2182,7 +2227,7 @@ <b>13.4.1.</b> - Top-k join processing + Top-k Join Processing </a> @@ -2197,7 +2242,7 @@ <b>13.4.2.</b> - Other utility functions + Other Utility Functions </a> @@ -2284,7 +2329,7 @@ <!-- Title --> <h1> <i class="fa fa-circle-o-notch fa-spin"></i> - <a href="../.." >E2006-tfidf regression tutorial for DataFrame</a> + <a href="../.." >E2006-tfidf Regression Tutorial for DataFrame</a> </h1> </div> @@ -2441,7 +2486,7 @@ Apache Hivemall is an effort undergoing incubation at The Apache Software Founda <script> var gitbook = gitbook || []; gitbook.push(function() { - gitbook.page.hasChanged({"page":{"title":"E2006-tfidf regression tutorial for DataFrame","level":"13.3.1","depth":2,"next":{"title":"E2006-tfidf regression tutorial for SQL","level":"13.3.2","depth":2,"path":"spark/regression/e2006_sql.md","ref":"spark/regression/e2006_sql.md","articles":[]},"previous":{"title":"Regression","level":"13.3","depth":1,"path":"spark/binaryclass/index.md","ref":"spark/binaryclass/index.md","articles":[{"title":"E2006-tfidf regression tutorial for DataFrame","level":"13.3.1","depth":2,"path":"spark/regression/e2006_df.md","ref":"spark/regression/e2006_df.md","articles":[]},{"title":"E2006-tfidf regression tutorial for SQL","level":"13.3.2","depth":2,"path":"spark/regression/e2006_sql.md","ref":"spark/regression/e2006_sql.md","articles":[]}]},"dir":"ltr"},"config":{"plugins":["theme-api","edit-link","github","splitter","sitemap","etoc","callouts","toggle-chapters","anchorjs","codeblock-filename","expandable-chapters","multipart","codeblock-file name","katex","emphasize","localized-footer"],"styles":{"website":"styles/website.css","pdf":"styles/pdf.css","epub":"styles/epub.css","mobi":"styles/mobi.css","ebook":"styles/ebook.css","print":"styles/print.css"},"pluginsConfig":{"emphasize":{},"callouts":{},"etoc":{"h2lb":3,"header":1,"maxdepth":3,"mindepth":1,"notoc":true},"github":{"url":"https://github.com/apache/incubator-hivemall/"},"splitter":{},"search":{},"downloadpdf":{"base":"https://github.com/apache/incubator-hivemall/docs/gitbook","label":"PDF","multilingual":false},"multipart":{},"localized-footer":{"filename":"FOOTER.md","hline":"true"},"lunr":{"maxIndexSize":1000000,"ignoreSpecialCharacters":false},"katex":{},"fontsettings":{"theme":"white","family":"sans","size":2,"font":"sans"},"highlight":{},"codeblock-filename":{},"sitemap":{"hostname":"https://hivemall.incubator.apache.org/"},"theme-api":{"languages":[],"split":false,"theme":"dark"},"sharing":{"facebook":true,"twitter":true,"google":false,"weibo":false,"insta paper":false,"vk":false,"all":["facebook","google","twitter","weibo","instapaper"]},"edit-link":{"label":"Edit","base":"https://github.com/apache/incubator-hivemall/tree/master/docs/gitbook"},"theme-default":{"styles":{"website":"styles/website.css","pdf":"styles/pdf.css","epub":"styles/epub.css","mobi":"styles/mobi.css","ebook":"styles/ebook.css","print":"styles/print.css"},"showLevel":true},"anchorjs":{"selector":"h1,h2,h3,*:not(.callout) > h4,h5"},"toggle-chapters":{},"expandable-chapters":{}},"theme":"default","pdf":{"pageNumbers":true,"fontSize":12,"fontFamily":"Arial","paperSize":"a4","chapterMark":"pagebreak","pageBreaksBefore":"/","margin":{"right":62,"left":62,"top":56,"bottom":56}},"structure":{"langs":"LANGS.md","readme":"README.md","glossary":"GLOSSARY.md","summary":"SUMMARY.md"},"variables":{},"title":"Hivemall User Manual","links":{"sidebar":{"<i class=\"fa fa-home\"></i> Home":"https://hivemall.incubator.apache.org/"}},"gitbook":"3.x.x","description":"User Manual for Apache Hivemall"},"file":{"path":"spark/regression/e2006_df.md","mtime":"2018-11-02T10:33:52.982Z","type":"markdown"},"gitbook":{"version":"3.2.3","time":"2018-11-13T09:32:29.643Z"},"basePath":"../..","book":{"language":""}}); + gitbook.page.hasChanged({"page":{"title":"E2006-tfidf Regression Tutorial for DataFrame","level":"13.3.1","depth":2,"next":{"title":"E2006-tfidf Regression Tutorial for SQL","level":"13.3.2","depth":2,"path":"spark/regression/e2006_sql.md","ref":"spark/regression/e2006_sql.md","articles":[]},"previous":{"title":"Regression","level":"13.3","depth":1,"path":"spark/binaryclass/index.md","ref":"spark/binaryclass/index.md","articles":[{"title":"E2006-tfidf Regression Tutorial for DataFrame","level":"13.3.1","depth":2,"path":"spark/regression/e2006_df.md","ref":"spark/regression/e2006_df.md","articles":[]},{"title":"E2006-tfidf Regression Tutorial for SQL","level":"13.3.2","depth":2,"path":"spark/regression/e2006_sql.md","ref":"spark/regression/e2006_sql.md","articles":[]}]},"dir":"ltr"},"config":{"plugins":["theme-api","edit-link","github","splitter","sitemap","etoc","callouts","toggle-chapters","anchorjs","codeblock-filename","expandable-chapters","multipart","codeblock-file name","katex","emphasize","localized-footer"],"styles":{"website":"styles/website.css","pdf":"styles/pdf.css","epub":"styles/epub.css","mobi":"styles/mobi.css","ebook":"styles/ebook.css","print":"styles/print.css"},"pluginsConfig":{"emphasize":{},"callouts":{},"etoc":{"h2lb":3,"header":1,"maxdepth":3,"mindepth":1,"notoc":true},"github":{"url":"https://github.com/apache/incubator-hivemall/"},"splitter":{},"search":{},"downloadpdf":{"base":"https://github.com/apache/incubator-hivemall/docs/gitbook","label":"PDF","multilingual":false},"multipart":{},"localized-footer":{"filename":"FOOTER.md","hline":"true"},"lunr":{"maxIndexSize":1000000,"ignoreSpecialCharacters":false},"katex":{},"fontsettings":{"theme":"white","family":"sans","size":2,"font":"sans"},"highlight":{},"codeblock-filename":{},"sitemap":{"hostname":"https://hivemall.incubator.apache.org/"},"theme-api":{"languages":[],"split":false,"theme":"dark"},"sharing":{"facebook":true,"twitter":true,"google":false,"weibo":false,"insta paper":false,"vk":false,"all":["facebook","google","twitter","weibo","instapaper"]},"edit-link":{"label":"Edit","base":"https://github.com/apache/incubator-hivemall/tree/master/docs/gitbook"},"theme-default":{"styles":{"website":"styles/website.css","pdf":"styles/pdf.css","epub":"styles/epub.css","mobi":"styles/mobi.css","ebook":"styles/ebook.css","print":"styles/print.css"},"showLevel":true},"anchorjs":{"selector":"h1,h2,h3,*:not(.callout) > h4,h5"},"toggle-chapters":{},"expandable-chapters":{}},"theme":"default","pdf":{"pageNumbers":true,"fontSize":12,"fontFamily":"Arial","paperSize":"a4","chapterMark":"pagebreak","pageBreaksBefore":"/","margin":{"right":62,"left":62,"top":56,"bottom":56}},"structure":{"langs":"LANGS.md","readme":"README.md","glossary":"GLOSSARY.md","summary":"SUMMARY.md"},"variables":{},"title":"Hivemall User Manual","links":{"sidebar":{"<i class=\"fa fa-home\"></i> Home":"https://hivemall.incubator.apache.org/"}},"gitbook":"3.x.x","description":"User Manual for Apache Hivemall"},"file":{"path":"spark/regression/e2006_df.md","mtime":"2018-12-25T07:50:37.551Z","type":"markdown"},"gitbook":{"version":"3.2.3","time":"2018-12-26T10:20:07.153Z"},"basePath":"../..","book":{"language":""}}); }); </script> </div> @@ -2471,7 +2516,7 @@ Apache Hivemall is an effort undergoing incubation at The Apache Software Founda - <script src="https://cdnjs.cloudflare.com/ajax/libs/anchor-js/4.1.1/anchor.min.js";></script> + <script src="../../gitbook/gitbook-plugin-anchorjs/anchor.min.js"></script>
http://git-wip-us.apache.org/repos/asf/incubator-hivemall-site/blob/d9012d92/userguide/spark/regression/e2006_sql.html ---------------------------------------------------------------------- diff --git a/userguide/spark/regression/e2006_sql.html b/userguide/spark/regression/e2006_sql.html index ec27d25..e310291 100644 --- a/userguide/spark/regression/e2006_sql.html +++ b/userguide/spark/regression/e2006_sql.html @@ -4,7 +4,7 @@ <head> <meta charset="UTF-8"> <meta content="text/html; charset=utf-8" http-equiv="Content-Type"> - <title>E2006-tfidf regression tutorial for SQL · Hivemall User Manual</title> + <title>E2006-tfidf Regression Tutorial for SQL · Hivemall User Manual</title> <meta http-equiv="X-UA-Compatible" content="IE=edge" /> <meta name="description" content=""> <meta name="generator" content="GitBook 3.2.3"> @@ -972,7 +972,7 @@ <b>6.2.1.</b> - Data preparation + Data Preparation </a> @@ -980,13 +980,28 @@ </li> - <li class="chapter " data-level="6.2.2" data-path="../../binaryclass/a9a_lr.html"> + <li class="chapter " data-level="6.2.2" data-path="../../binaryclass/a9a_generic.html"> - <a href="../../binaryclass/a9a_lr.html"> + <a href="../../binaryclass/a9a_generic.html"> <b>6.2.2.</b> + General Binary Classifier + + </a> + + + + </li> + + <li class="chapter " data-level="6.2.3" data-path="../../binaryclass/a9a_lr.html"> + + <a href="../../binaryclass/a9a_lr.html"> + + + <b>6.2.3.</b> + Logistic Regression </a> @@ -995,14 +1010,14 @@ </li> - <li class="chapter " data-level="6.2.3" data-path="../../binaryclass/a9a_minibatch.html"> + <li class="chapter " data-level="6.2.4" data-path="../../binaryclass/a9a_minibatch.html"> <a href="../../binaryclass/a9a_minibatch.html"> - <b>6.2.3.</b> + <b>6.2.4.</b> - Mini-batch gradient descent + Mini-batch Gradient Descent </a> @@ -1038,7 +1053,7 @@ <b>6.3.1.</b> - Data preparation + Data Preparation </a> @@ -1076,13 +1091,28 @@ </li> - <li class="chapter " data-level="6.3.4" data-path="../../binaryclass/news20_adagrad.html"> + <li class="chapter " data-level="6.3.4" data-path="../../binaryclass/news20_generic.html"> - <a href="../../binaryclass/news20_adagrad.html"> + <a href="../../binaryclass/news20_generic.html"> <b>6.3.4.</b> + General Binary Classifier + + </a> + + + + </li> + + <li class="chapter " data-level="6.3.5" data-path="../../binaryclass/news20_adagrad.html"> + + <a href="../../binaryclass/news20_adagrad.html"> + + + <b>6.3.5.</b> + AdaGradRDA, AdaGrad, AdaDelta </a> @@ -1091,12 +1121,12 @@ </li> - <li class="chapter " data-level="6.3.5" data-path="../../binaryclass/news20_rf.html"> + <li class="chapter " data-level="6.3.6" data-path="../../binaryclass/news20_rf.html"> <a href="../../binaryclass/news20_rf.html"> - <b>6.3.5.</b> + <b>6.3.6.</b> Random Forest @@ -1134,7 +1164,7 @@ <b>6.4.1.</b> - Data preparation + Data Preparation </a> @@ -1185,7 +1215,7 @@ <b>6.5.1.</b> - Data preparation + Data Preparation </a> @@ -1236,7 +1266,7 @@ <b>6.6.1.</b> - Data pareparation + Data Pareparation </a> @@ -1302,7 +1332,7 @@ <b>6.8.1.</b> - Data preparation + Data Preparation </a> @@ -1360,7 +1390,7 @@ <b>7.1.1.</b> - Data preparation + Data Preparation </a> @@ -1375,7 +1405,7 @@ <b>7.1.2.</b> - Data preparation for one-vs-the-rest classifiers + Data Preparation for one-vs-the-rest classifiers </a> @@ -1435,7 +1465,7 @@ <b>7.1.6.</b> - one-vs-the-rest classifier + one-vs-the-rest Classifier </a> @@ -1559,7 +1589,7 @@ <b>8.2.1.</b> - Data preparation + Data Preparation </a> @@ -1567,13 +1597,28 @@ </li> - <li class="chapter " data-level="8.2.2" data-path="../../regression/e2006_arow.html"> + <li class="chapter " data-level="8.2.2" data-path="../../regression/e2006_generic.html"> - <a href="../../regression/e2006_arow.html"> + <a href="../../regression/e2006_generic.html"> <b>8.2.2.</b> + General Regessor + + </a> + + + + </li> + + <li class="chapter " data-level="8.2.3" data-path="../../regression/e2006_arow.html"> + + <a href="../../regression/e2006_arow.html"> + + + <b>8.2.3.</b> + Passive Aggressive, AROW </a> @@ -1610,7 +1655,7 @@ <b>8.3.1.</b> - Data preparation + Data Preparation </a> @@ -1698,7 +1743,7 @@ <b>9.1.1.</b> - Item-based collaborative filtering + Item-based Collaborative Filtering </a> @@ -1734,7 +1779,7 @@ <b>9.2.1.</b> - Data preparation + Data Preparation </a> @@ -1749,7 +1794,7 @@ <b>9.2.2.</b> - LSH/MinHash and Jaccard similarity + LSH/MinHash and Jaccard Similarity </a> @@ -1764,7 +1809,7 @@ <b>9.2.3.</b> - LSH/MinHash and brute-force search + LSH/MinHash and Brute-force Search </a> @@ -1815,7 +1860,7 @@ <b>9.3.1.</b> - Data preparation + Data Preparation </a> @@ -1830,7 +1875,7 @@ <b>9.3.2.</b> - Item-based collaborative filtering + Item-based Collaborative Filtering </a> @@ -1875,7 +1920,7 @@ <b>9.3.5.</b> - SLIM for fast top-k recommendation + SLIM for fast top-k Recommendation </a> @@ -1890,7 +1935,7 @@ <b>9.3.6.</b> - 10-fold cross validation (Matrix Factorization) + 10-fold Cross Validation (Matrix Factorization) </a> @@ -2080,7 +2125,7 @@ <b>13.2.1.</b> - a9a tutorial for DataFrame + a9a Tutorial for DataFrame </a> @@ -2095,7 +2140,7 @@ <b>13.2.2.</b> - a9a tutorial for SQL + a9a Tutorial for SQL </a> @@ -2131,7 +2176,7 @@ <b>13.3.1.</b> - E2006-tfidf regression tutorial for DataFrame + E2006-tfidf Regression Tutorial for DataFrame </a> @@ -2146,7 +2191,7 @@ <b>13.3.2.</b> - E2006-tfidf regression tutorial for SQL + E2006-tfidf Regression Tutorial for SQL </a> @@ -2166,7 +2211,7 @@ <b>13.4.</b> - Generic features + Generic Features </a> @@ -2182,7 +2227,7 @@ <b>13.4.1.</b> - Top-k join processing + Top-k Join Processing </a> @@ -2197,7 +2242,7 @@ <b>13.4.2.</b> - Other utility functions + Other Utility Functions </a> @@ -2284,7 +2329,7 @@ <!-- Title --> <h1> <i class="fa fa-circle-o-notch fa-spin"></i> - <a href="../.." >E2006-tfidf regression tutorial for SQL</a> + <a href="../.." >E2006-tfidf Regression Tutorial for SQL</a> </h1> </div> @@ -2487,7 +2532,7 @@ Apache Hivemall is an effort undergoing incubation at The Apache Software Founda <script> var gitbook = gitbook || []; gitbook.push(function() { - gitbook.page.hasChanged({"page":{"title":"E2006-tfidf regression tutorial for SQL","level":"13.3.2","depth":2,"next":{"title":"Generic features","level":"13.4","depth":1,"path":"spark/misc/misc.md","ref":"spark/misc/misc.md","articles":[{"title":"Top-k join processing","level":"13.4.1","depth":2,"path":"spark/misc/topk_join.md","ref":"spark/misc/topk_join.md","articles":[]},{"title":"Other utility functions","level":"13.4.2","depth":2,"path":"spark/misc/functions.md","ref":"spark/misc/functions.md","articles":[]}]},"previous":{"title":"E2006-tfidf regression tutorial for DataFrame","level":"13.3.1","depth":2,"path":"spark/regression/e2006_df.md","ref":"spark/regression/e2006_df.md","articles":[]},"dir":"ltr"},"config":{"plugins":["theme-api","edit-link","github","splitter","sitemap","etoc","callouts","toggle-chapters","anchorjs","codeblock-filename","expandable-chapters","multipart","codeblock-filename","katex","emphasize","localized-footer"],"styles":{"website":"styles/ website.css","pdf":"styles/pdf.css","epub":"styles/epub.css","mobi":"styles/mobi.css","ebook":"styles/ebook.css","print":"styles/print.css"},"pluginsConfig":{"emphasize":{},"callouts":{},"etoc":{"h2lb":3,"header":1,"maxdepth":3,"mindepth":1,"notoc":true},"github":{"url":"https://github.com/apache/incubator-hivemall/"},"splitter":{},"search":{},"downloadpdf":{"base":"https://github.com/apache/incubator-hivemall/docs/gitbook","label":"PDF","multilingual":false},"multipart":{},"localized-footer":{"filename":"FOOTER.md","hline":"true"},"lunr":{"maxIndexSize":1000000,"ignoreSpecialCharacters":false},"katex":{},"fontsettings":{"theme":"white","family":"sans","size":2,"font":"sans"},"highlight":{},"codeblock-filename":{},"sitemap":{"hostname":"https://hivemall.incubator.apache.org/"},"theme-api":{"languages":[],"split":false,"theme":"dark"},"sharing":{"facebook":true,"twitter":true,"google":false,"weibo":false,"instapaper":false,"vk":false,"all":["facebook","google","twitter","weibo","inst apaper"]},"edit-link":{"label":"Edit","base":"https://github.com/apache/incubator-hivemall/tree/master/docs/gitbook"},"theme-default":{"styles":{"website":"styles/website.css","pdf":"styles/pdf.css","epub":"styles/epub.css","mobi":"styles/mobi.css","ebook":"styles/ebook.css","print":"styles/print.css"},"showLevel":true},"anchorjs":{"selector":"h1,h2,h3,*:not(.callout) > h4,h5"},"toggle-chapters":{},"expandable-chapters":{}},"theme":"default","pdf":{"pageNumbers":true,"fontSize":12,"fontFamily":"Arial","paperSize":"a4","chapterMark":"pagebreak","pageBreaksBefore":"/","margin":{"right":62,"left":62,"top":56,"bottom":56}},"structure":{"langs":"LANGS.md","readme":"README.md","glossary":"GLOSSARY.md","summary":"SUMMARY.md"},"variables":{},"title":"Hivemall User Manual","links":{"sidebar":{"<i class=\"fa fa-home\"></i> Home":"https://hivemall.incubator.apache.org/"}},"gitbook":"3.x.x","description":"User Manual for Apache Hivemall"},"file":{"path":"spark/regression/e2006_sql.md","mtime":" 2018-11-02T10:33:52.983Z","type":"markdown"},"gitbook":{"version":"3.2.3","time":"2018-11-13T09:32:29.643Z"},"basePath":"../..","book":{"language":""}}); + gitbook.page.hasChanged({"page":{"title":"E2006-tfidf Regression Tutorial for SQL","level":"13.3.2","depth":2,"next":{"title":"Generic Features","level":"13.4","depth":1,"path":"spark/misc/misc.md","ref":"spark/misc/misc.md","articles":[{"title":"Top-k Join Processing","level":"13.4.1","depth":2,"path":"spark/misc/topk_join.md","ref":"spark/misc/topk_join.md","articles":[]},{"title":"Other Utility Functions","level":"13.4.2","depth":2,"path":"spark/misc/functions.md","ref":"spark/misc/functions.md","articles":[]}]},"previous":{"title":"E2006-tfidf Regression Tutorial for DataFrame","level":"13.3.1","depth":2,"path":"spark/regression/e2006_df.md","ref":"spark/regression/e2006_df.md","articles":[]},"dir":"ltr"},"config":{"plugins":["theme-api","edit-link","github","splitter","sitemap","etoc","callouts","toggle-chapters","anchorjs","codeblock-filename","expandable-chapters","multipart","codeblock-filename","katex","emphasize","localized-footer"],"styles":{"website":"styles/ website.css","pdf":"styles/pdf.css","epub":"styles/epub.css","mobi":"styles/mobi.css","ebook":"styles/ebook.css","print":"styles/print.css"},"pluginsConfig":{"emphasize":{},"callouts":{},"etoc":{"h2lb":3,"header":1,"maxdepth":3,"mindepth":1,"notoc":true},"github":{"url":"https://github.com/apache/incubator-hivemall/"},"splitter":{},"search":{},"downloadpdf":{"base":"https://github.com/apache/incubator-hivemall/docs/gitbook","label":"PDF","multilingual":false},"multipart":{},"localized-footer":{"filename":"FOOTER.md","hline":"true"},"lunr":{"maxIndexSize":1000000,"ignoreSpecialCharacters":false},"katex":{},"fontsettings":{"theme":"white","family":"sans","size":2,"font":"sans"},"highlight":{},"codeblock-filename":{},"sitemap":{"hostname":"https://hivemall.incubator.apache.org/"},"theme-api":{"languages":[],"split":false,"theme":"dark"},"sharing":{"facebook":true,"twitter":true,"google":false,"weibo":false,"instapaper":false,"vk":false,"all":["facebook","google","twitter","weibo","inst apaper"]},"edit-link":{"label":"Edit","base":"https://github.com/apache/incubator-hivemall/tree/master/docs/gitbook"},"theme-default":{"styles":{"website":"styles/website.css","pdf":"styles/pdf.css","epub":"styles/epub.css","mobi":"styles/mobi.css","ebook":"styles/ebook.css","print":"styles/print.css"},"showLevel":true},"anchorjs":{"selector":"h1,h2,h3,*:not(.callout) > h4,h5"},"toggle-chapters":{},"expandable-chapters":{}},"theme":"default","pdf":{"pageNumbers":true,"fontSize":12,"fontFamily":"Arial","paperSize":"a4","chapterMark":"pagebreak","pageBreaksBefore":"/","margin":{"right":62,"left":62,"top":56,"bottom":56}},"structure":{"langs":"LANGS.md","readme":"README.md","glossary":"GLOSSARY.md","summary":"SUMMARY.md"},"variables":{},"title":"Hivemall User Manual","links":{"sidebar":{"<i class=\"fa fa-home\"></i> Home":"https://hivemall.incubator.apache.org/"}},"gitbook":"3.x.x","description":"User Manual for Apache Hivemall"},"file":{"path":"spark/regression/e2006_sql.md","mtime":" 2018-12-25T07:50:37.551Z","type":"markdown"},"gitbook":{"version":"3.2.3","time":"2018-12-26T10:20:07.153Z"},"basePath":"../..","book":{"language":""}}); }); </script> </div> @@ -2517,7 +2562,7 @@ Apache Hivemall is an effort undergoing incubation at The Apache Software Founda - <script src="https://cdnjs.cloudflare.com/ajax/libs/anchor-js/4.1.1/anchor.min.js";></script> + <script src="../../gitbook/gitbook-plugin-anchorjs/anchor.min.js"></script> http://git-wip-us.apache.org/repos/asf/incubator-hivemall-site/blob/d9012d92/userguide/supervised_learning/prediction.html ---------------------------------------------------------------------- diff --git a/userguide/supervised_learning/prediction.html b/userguide/supervised_learning/prediction.html index 24d1c45..0b86e68 100644 --- a/userguide/supervised_learning/prediction.html +++ b/userguide/supervised_learning/prediction.html @@ -972,7 +972,7 @@ <b>6.2.1.</b> - Data preparation + Data Preparation </a> @@ -980,13 +980,28 @@ </li> - <li class="chapter " data-level="6.2.2" data-path="../binaryclass/a9a_lr.html"> + <li class="chapter " data-level="6.2.2" data-path="../binaryclass/a9a_generic.html"> - <a href="../binaryclass/a9a_lr.html"> + <a href="../binaryclass/a9a_generic.html"> <b>6.2.2.</b> + General Binary Classifier + + </a> + + + + </li> + + <li class="chapter " data-level="6.2.3" data-path="../binaryclass/a9a_lr.html"> + + <a href="../binaryclass/a9a_lr.html"> + + + <b>6.2.3.</b> + Logistic Regression </a> @@ -995,14 +1010,14 @@ </li> - <li class="chapter " data-level="6.2.3" data-path="../binaryclass/a9a_minibatch.html"> + <li class="chapter " data-level="6.2.4" data-path="../binaryclass/a9a_minibatch.html"> <a href="../binaryclass/a9a_minibatch.html"> - <b>6.2.3.</b> + <b>6.2.4.</b> - Mini-batch gradient descent + Mini-batch Gradient Descent </a> @@ -1038,7 +1053,7 @@ <b>6.3.1.</b> - Data preparation + Data Preparation </a> @@ -1076,13 +1091,28 @@ </li> - <li class="chapter " data-level="6.3.4" data-path="../binaryclass/news20_adagrad.html"> + <li class="chapter " data-level="6.3.4" data-path="../binaryclass/news20_generic.html"> - <a href="../binaryclass/news20_adagrad.html"> + <a href="../binaryclass/news20_generic.html"> <b>6.3.4.</b> + General Binary Classifier + + </a> + + + + </li> + + <li class="chapter " data-level="6.3.5" data-path="../binaryclass/news20_adagrad.html"> + + <a href="../binaryclass/news20_adagrad.html"> + + + <b>6.3.5.</b> + AdaGradRDA, AdaGrad, AdaDelta </a> @@ -1091,12 +1121,12 @@ </li> - <li class="chapter " data-level="6.3.5" data-path="../binaryclass/news20_rf.html"> + <li class="chapter " data-level="6.3.6" data-path="../binaryclass/news20_rf.html"> <a href="../binaryclass/news20_rf.html"> - <b>6.3.5.</b> + <b>6.3.6.</b> Random Forest @@ -1134,7 +1164,7 @@ <b>6.4.1.</b> - Data preparation + Data Preparation </a> @@ -1185,7 +1215,7 @@ <b>6.5.1.</b> - Data preparation + Data Preparation </a> @@ -1236,7 +1266,7 @@ <b>6.6.1.</b> - Data pareparation + Data Pareparation </a> @@ -1302,7 +1332,7 @@ <b>6.8.1.</b> - Data preparation + Data Preparation </a> @@ -1360,7 +1390,7 @@ <b>7.1.1.</b> - Data preparation + Data Preparation </a> @@ -1375,7 +1405,7 @@ <b>7.1.2.</b> - Data preparation for one-vs-the-rest classifiers + Data Preparation for one-vs-the-rest classifiers </a> @@ -1435,7 +1465,7 @@ <b>7.1.6.</b> - one-vs-the-rest classifier + one-vs-the-rest Classifier </a> @@ -1559,7 +1589,7 @@ <b>8.2.1.</b> - Data preparation + Data Preparation </a> @@ -1567,13 +1597,28 @@ </li> - <li class="chapter " data-level="8.2.2" data-path="../regression/e2006_arow.html"> + <li class="chapter " data-level="8.2.2" data-path="../regression/e2006_generic.html"> - <a href="../regression/e2006_arow.html"> + <a href="../regression/e2006_generic.html"> <b>8.2.2.</b> + General Regessor + + </a> + + + + </li> + + <li class="chapter " data-level="8.2.3" data-path="../regression/e2006_arow.html"> + + <a href="../regression/e2006_arow.html"> + + + <b>8.2.3.</b> + Passive Aggressive, AROW </a> @@ -1610,7 +1655,7 @@ <b>8.3.1.</b> - Data preparation + Data Preparation </a> @@ -1698,7 +1743,7 @@ <b>9.1.1.</b> - Item-based collaborative filtering + Item-based Collaborative Filtering </a> @@ -1734,7 +1779,7 @@ <b>9.2.1.</b> - Data preparation + Data Preparation </a> @@ -1749,7 +1794,7 @@ <b>9.2.2.</b> - LSH/MinHash and Jaccard similarity + LSH/MinHash and Jaccard Similarity </a> @@ -1764,7 +1809,7 @@ <b>9.2.3.</b> - LSH/MinHash and brute-force search + LSH/MinHash and Brute-force Search </a> @@ -1815,7 +1860,7 @@ <b>9.3.1.</b> - Data preparation + Data Preparation </a> @@ -1830,7 +1875,7 @@ <b>9.3.2.</b> - Item-based collaborative filtering + Item-based Collaborative Filtering </a> @@ -1875,7 +1920,7 @@ <b>9.3.5.</b> - SLIM for fast top-k recommendation + SLIM for fast top-k Recommendation </a> @@ -1890,7 +1935,7 @@ <b>9.3.6.</b> - 10-fold cross validation (Matrix Factorization) + 10-fold Cross Validation (Matrix Factorization) </a> @@ -2080,7 +2125,7 @@ <b>13.2.1.</b> - a9a tutorial for DataFrame + a9a Tutorial for DataFrame </a> @@ -2095,7 +2140,7 @@ <b>13.2.2.</b> - a9a tutorial for SQL + a9a Tutorial for SQL </a> @@ -2131,7 +2176,7 @@ <b>13.3.1.</b> - E2006-tfidf regression tutorial for DataFrame + E2006-tfidf Regression Tutorial for DataFrame </a> @@ -2146,7 +2191,7 @@ <b>13.3.2.</b> - E2006-tfidf regression tutorial for SQL + E2006-tfidf Regression Tutorial for SQL </a> @@ -2166,7 +2211,7 @@ <b>13.4.</b> - Generic features + Generic Features </a> @@ -2182,7 +2227,7 @@ <b>13.4.1.</b> - Top-k join processing + Top-k Join Processing </a> @@ -2197,7 +2242,7 @@ <b>13.4.2.</b> - Other utility functions + Other Utility Functions </a> @@ -2481,12 +2526,122 @@ E(\mathbf{w}) := \frac{1}{n} \sum_{i=1}^{n} L(\mathbf{w}; \mathbf{x}_i, y_i) + \ <ul> <li>Optimizer: <code>-opt</code>, <code>-optimizer</code><ul> <li>SGD</li> -<li>AdaGrad</li> -<li>AdaDelta</li> -<li>Adam</li> +<li>Momentum<ul> +<li>Hyperparameters<ul> +<li><code>-alpha 1.0</code> Learning rate.</li> +<li><code>-momentum 0.9</code> Exponential decay rate of the first order moment.</li> +</ul> +</li> +</ul> +</li> +<li>Nesterov<ul> +<li>See: <a href="https://arxiv.org/abs/1212.0901"; target="_blank">https://arxiv.org/abs/1212.0901</a></li> +<li>Hyperparameters<ul> +<li>same as Momentum</li> +</ul> +</li> +</ul> +</li> +<li>AdaGrad (default)<ul> +<li>See: <a href="http://jmlr.org/papers/v12/duchi11a.html"; target="_blank">http://jmlr.org/papers/v12/duchi11a.html</a></li> +<li>Hyperparameters<ul> +<li><code>-eps 1.0</code> Constant for the numerical stability.</li> +</ul> +</li> +</ul> +</li> +<li>RMSprop<ul> +<li>Description: RMSprop optimizer introducing weight decay to AdaGrad.</li> +<li>See: <a href="http://www.cs.toronto.edu/~tijmen/csc321/slides/lecture_slides_lec6.pdf"; target="_blank">http://www.cs.toronto.edu/~tijmen/csc321/slides/lecture_slides_lec6.pdf</a></li> +<li>Hyperparameters<ul> +<li><code>-decay 0.95</code> Weight decay rate</li> +<li><code>-eps 1.0</code> Constant for numerical stability</li> +</ul> +</li> +</ul> +</li> +<li>RMSpropGraves<ul> +<li>Description: Alex Graves's RMSprop introducing weight decay and momentum.</li> +<li>See: <a href="https://arxiv.org/abs/1308.0850"; target="_blank">https://arxiv.org/abs/1308.0850</a></li> +<li>Hyperparameters<ul> +<li><code>-alpha 1.0</code> Learning rate.</li> +<li><code>-decay 0.95</code> Weight decay rate</li> +<li><code>-momentum 0.9</code> Exponential decay rate of the first order moment.</li> +<li><code>-eps 1.0</code> Constant for numerical stability</li> +</ul> +</li> +</ul> +</li> +<li>AdaDelta<ul> +<li>See: <a href="https://arxiv.org/abs/1212.5701"; target="_blank">https://arxiv.org/abs/1212.5701</a></li> +<li>Hyperparameters<ul> +<li><code>-decay 0.95</code> Weight decay rate</li> +<li><code>-eps 1e-6f</code> Constant for numerical stability</li> +</ul> +</li> </ul> </li> +<li>Adam<ul> +<li>See:<ul> +<li><a href="https://arxiv.org/abs/1412.6980v8"; target="_blank">Adam: A Method for Stochastic Optimization</a></li> +<li><a href="https://openreview.net/forum?id=rk6qdGgCZ"; target="_blank">Fixing Weight Decay Regularization in Adam</a></li> +<li><a href="https://openreview.net/forum?id=ryQu7f-RZ"; target="_blank">On the Convergence of Adam and Beyond</a></li> </ul> +</li> +<li>Hyperparameters<ul> +<li><code>-alpha 1.0</code> Learning rate.</li> +<li><code>-beta1 0.9</code> Exponential decay rate of the first order moment.</li> +<li><code>-beta2 0.999</code> Exponential decay rate of the second order moment.</li> +<li><code>-eps 1e-8f</code> Constant for numerical stability</li> +<li><code>-decay 0.0</code> Weight decay rate</li> +</ul> +</li> +</ul> +</li> +<li>Nadam<ul> +<li>Description: Nadam is Adam optimizer with Nesterov momentum.</li> +<li>See:<ul> +<li><a href="https://openreview.net/pdf?id=OM0jvwB8jIp57ZJjtNEZ"; target="_blank">Incorporating Nesterov Momentum into Adam</a></li> +<li><a href="http://cs229.stanford.edu/proj2015/054_report.pdf"; target="_blank">Adam report</a></li> +<li><a href="http://www.cs.toronto.edu/~fritz/absps/momentum.pdf"; target="_blank">On the importance of initialization and momentum in deep learning</a></li> +</ul> +</li> +<li>Hyperparameters<ul> +<li>same as Adam except ...</li> +<li><code>-scheduleDecay 0.004</code> Scheduled decay rate (for each 250 steps by the default; 1/250=0.004)</li> +</ul> +</li> +</ul> +</li> +<li>Eve<ul> +<li>See: <a href="https://openreview.net/forum?id=r1WUqIceg"; target="_blank">https://openreview.net/forum?id=r1WUqIceg</a></li> +<li>Hyperparameters<ul> +<li>same as Adam except ...</li> +<li><code>-beta3 0.999</code> Decay rate for Eve coefficient.</li> +<li><code>-c 10</code> Constant used for gradient clipping <code>clip(val, 1/c, c)</code></li> +</ul> +</li> +</ul> +</li> +<li>AdamHD<ul> +<li>Description: Adam optimizer with Hypergradient Descent. Learning rate <code>-alpha</code> is automatically tuned.</li> +<li>See:<ul> +<li><a href="https://openreview.net/forum?id=BkrsAzWAb"; target="_blank">Online Learning Rate Adaptation with Hypergradient Descent</a></li> +<li><a href="https://damaru2.github.io/convergence_analysis_hypergradient_descent/dissertation_hypergradients.pdf"; target="_blank">Convergence Analysis of an Adaptive Method of Gradient Descent</a></li> +</ul> +</li> +<li>Hyperparameters<ul> +<li>same as Adam except ...</li> +<li><code>-alpha 0.02</code> Learning rate.</li> +<li><code>-beta -1e-6</code> Constant used for tuning learning rate.</li> +</ul> +</li> +</ul> +</li> +</ul> +</li> +</ul> +<p>Default (Adagrad+RDA), AdaDelta, Adam, and AdamHD is worth trying in my experience.</p> <div class="panel panel-primary"><div class="panel-heading"><h3 class="panel-title" id="note"><i class="fa fa-edit"></i> Note</h3></div><div class="panel-body"><p>Option values are case insensitive and you can use <code>sgd</code> or <code>rda</code>, or <code>huberloss</code> in lower-case letters.</p></div></div> <p>Furthermore, optimizer offers to set auxiliary options such as:</p> <ul> @@ -2506,8 +2661,13 @@ E(\mathbf{w}) := \frac{1}{n} \sum_{i=1}^{n} L(\mathbf{w}; \mathbf{x}_i, y_i) + \ </li> </ul> <p>For details of available options, following queries might be helpful to list all of them:</p> -<pre><code class="lang-sql"><span class="hljs-keyword">select</span> train_regressor(<span class="hljs-built_in">array</span>(), <span class="hljs-number">0</span>, <span class="hljs-string">'-help'</span>); -<span class="hljs-keyword">select</span> train_classifier(<span class="hljs-built_in">array</span>(), <span class="hljs-number">0</span>, <span class="hljs-string">'-help'</span>); +<pre><code class="lang-sql"><span class="hljs-keyword">select</span> train_regressor(<span class="hljs-string">'-help'</span>); +<span class="hljs-comment">-- v0.5.0 or before</span> +<span class="hljs-comment">-- select train_regressor(array(), 0, '-help');</span> + +<span class="hljs-keyword">select</span> train_classifier(<span class="hljs-string">'-help'</span>); +<span class="hljs-comment">-- v0.5.0 or before</span> +<span class="hljs-comment">-- select train_classifier(array(), 0, '-help');</span> </code></pre> <p>In practice, you can try different combinations of the options in order to achieve higher prediction accuracy.</p> <p><div id="page-footer" class="localized-footer"><hr><!-- @@ -2565,7 +2725,7 @@ Apache Hivemall is an effort undergoing incubation at The Apache Software Founda <script> var gitbook = gitbook || []; gitbook.push(function() { - gitbook.page.hasChanged({"page":{"title":"How Prediction Works","level":"5.1","depth":1,"next":{"title":"Step-by-Step Tutorial on Supervised Learning","level":"5.2","depth":1,"path":"supervised_learning/tutorial.md","ref":"supervised_learning/tutorial.md","articles":[]},"previous":{"title":"Logistic Regression data generation","level":"4.5.1","depth":2,"path":"eval/lr_datagen.md","ref":"eval/lr_datagen.md","articles":[]},"dir":"ltr"},"config":{"plugins":["theme-api","edit-link","github","splitter","sitemap","etoc","callouts","toggle-chapters","anchorjs","codeblock-filename","expandable-chapters","multipart","codeblock-filename","katex","emphasize","localized-footer"],"styles":{"website":"styles/website.css","pdf":"styles/pdf.css","epub":"styles/epub.css","mobi":"styles/mobi.css","ebook":"styles/ebook.css","print":"styles/print.css"},"pluginsConfig":{"emphasize":{},"callouts":{},"etoc":{"h2lb":3,"header":1,"maxdepth":3,"mindepth":1,"notoc":true},"github":{"url":"https://g ithub.com/apache/incubator-hivemall/"},"splitter":{},"search":{},"downloadpdf":{"base":"https://github.com/apache/incubator-hivemall/docs/gitbook","label":"PDF","multilingual":false},"multipart":{},"localized-footer":{"filename":"FOOTER.md","hline":"true"},"lunr":{"maxIndexSize":1000000,"ignoreSpecialCharacters":false},"katex":{},"fontsettings":{"theme":"white","family":"sans","size":2,"font":"sans"},"highlight":{},"codeblock-filename":{},"sitemap":{"hostname":"https://hivemall.incubator.apache.org/"},"theme-api":{"languages":[],"split":false,"theme":"dark"},"sharing":{"facebook":true,"twitter":true,"google":false,"weibo":false,"instapaper":false,"vk":false,"all":["facebook","google","twitter","weibo","instapaper"]},"edit-link":{"label":"Edit","base":"https://github.com/apache/incubator-hivemall/tree/master/docs/gitbook"},"theme-default":{"styles":{"website":"styles/website.css","pdf":"styles/pdf.css","epub":"styles/epub.css","mobi":"styles/mobi.css","ebook":"styles/ebook.css","prin t":"styles/print.css"},"showLevel":true},"anchorjs":{"selector":"h1,h2,h3,*:not(.callout) > h4,h5"},"toggle-chapters":{},"expandable-chapters":{}},"theme":"default","pdf":{"pageNumbers":true,"fontSize":12,"fontFamily":"Arial","paperSize":"a4","chapterMark":"pagebreak","pageBreaksBefore":"/","margin":{"right":62,"left":62,"top":56,"bottom":56}},"structure":{"langs":"LANGS.md","readme":"README.md","glossary":"GLOSSARY.md","summary":"SUMMARY.md"},"variables":{},"title":"Hivemall User Manual","links":{"sidebar":{"<i class=\"fa fa-home\"></i> Home":"https://hivemall.incubator.apache.org/"}},"gitbook":"3.x.x","description":"User Manual for Apache Hivemall"},"file":{"path":"supervised_learning/prediction.md","mtime":"2018-11-02T10:33:52.984Z","type":"markdown"},"gitbook":{"version":"3.2.3","time":"2018-11-13T09:32:29.643Z"},"basePath":"..","book":{"language":""}}); + gitbook.page.hasChanged({"page":{"title":"How Prediction Works","level":"5.1","depth":1,"next":{"title":"Step-by-Step Tutorial on Supervised Learning","level":"5.2","depth":1,"path":"supervised_learning/tutorial.md","ref":"supervised_learning/tutorial.md","articles":[]},"previous":{"title":"Logistic Regression data generation","level":"4.5.1","depth":2,"path":"eval/lr_datagen.md","ref":"eval/lr_datagen.md","articles":[]},"dir":"ltr"},"config":{"plugins":["theme-api","edit-link","github","splitter","sitemap","etoc","callouts","toggle-chapters","anchorjs","codeblock-filename","expandable-chapters","multipart","codeblock-filename","katex","emphasize","localized-footer"],"styles":{"website":"styles/website.css","pdf":"styles/pdf.css","epub":"styles/epub.css","mobi":"styles/mobi.css","ebook":"styles/ebook.css","print":"styles/print.css"},"pluginsConfig":{"emphasize":{},"callouts":{},"etoc":{"h2lb":3,"header":1,"maxdepth":3,"mindepth":1,"notoc":true},"github":{"url":"https://g ithub.com/apache/incubator-hivemall/"},"splitter":{},"search":{},"downloadpdf":{"base":"https://github.com/apache/incubator-hivemall/docs/gitbook","label":"PDF","multilingual":false},"multipart":{},"localized-footer":{"filename":"FOOTER.md","hline":"true"},"lunr":{"maxIndexSize":1000000,"ignoreSpecialCharacters":false},"katex":{},"fontsettings":{"theme":"white","family":"sans","size":2,"font":"sans"},"highlight":{},"codeblock-filename":{},"sitemap":{"hostname":"https://hivemall.incubator.apache.org/"},"theme-api":{"languages":[],"split":false,"theme":"dark"},"sharing":{"facebook":true,"twitter":true,"google":false,"weibo":false,"instapaper":false,"vk":false,"all":["facebook","google","twitter","weibo","instapaper"]},"edit-link":{"label":"Edit","base":"https://github.com/apache/incubator-hivemall/tree/master/docs/gitbook"},"theme-default":{"styles":{"website":"styles/website.css","pdf":"styles/pdf.css","epub":"styles/epub.css","mobi":"styles/mobi.css","ebook":"styles/ebook.css","prin t":"styles/print.css"},"showLevel":true},"anchorjs":{"selector":"h1,h2,h3,*:not(.callout) > h4,h5"},"toggle-chapters":{},"expandable-chapters":{}},"theme":"default","pdf":{"pageNumbers":true,"fontSize":12,"fontFamily":"Arial","paperSize":"a4","chapterMark":"pagebreak","pageBreaksBefore":"/","margin":{"right":62,"left":62,"top":56,"bottom":56}},"structure":{"langs":"LANGS.md","readme":"README.md","glossary":"GLOSSARY.md","summary":"SUMMARY.md"},"variables":{},"title":"Hivemall User Manual","links":{"sidebar":{"<i class=\"fa fa-home\"></i> Home":"https://hivemall.incubator.apache.org/"}},"gitbook":"3.x.x","description":"User Manual for Apache Hivemall"},"file":{"path":"supervised_learning/prediction.md","mtime":"2018-12-26T10:16:03.082Z","type":"markdown"},"gitbook":{"version":"3.2.3","time":"2018-12-26T10:20:07.153Z"},"basePath":"..","book":{"language":""}}); }); </script> </div> @@ -2595,7 +2755,7 @@ Apache Hivemall is an effort undergoing incubation at The Apache Software Founda - <script src="https://cdnjs.cloudflare.com/ajax/libs/anchor-js/4.1.1/anchor.min.js";></script> + <script src="../gitbook/gitbook-plugin-anchorjs/anchor.min.js"></script> http://git-wip-us.apache.org/repos/asf/incubator-hivemall-site/blob/d9012d92/userguide/supervised_learning/tutorial.html ---------------------------------------------------------------------- diff --git a/userguide/supervised_learning/tutorial.html b/userguide/supervised_learning/tutorial.html index 67e3a41..559f60a 100644 --- a/userguide/supervised_learning/tutorial.html +++ b/userguide/supervised_learning/tutorial.html @@ -972,7 +972,7 @@ <b>6.2.1.</b> - Data preparation + Data Preparation </a> @@ -980,13 +980,28 @@ </li> - <li class="chapter " data-level="6.2.2" data-path="../binaryclass/a9a_lr.html"> + <li class="chapter " data-level="6.2.2" data-path="../binaryclass/a9a_generic.html"> - <a href="../binaryclass/a9a_lr.html"> + <a href="../binaryclass/a9a_generic.html"> <b>6.2.2.</b> + General Binary Classifier + + </a> + + + + </li> + + <li class="chapter " data-level="6.2.3" data-path="../binaryclass/a9a_lr.html"> + + <a href="../binaryclass/a9a_lr.html"> + + + <b>6.2.3.</b> + Logistic Regression </a> @@ -995,14 +1010,14 @@ </li> - <li class="chapter " data-level="6.2.3" data-path="../binaryclass/a9a_minibatch.html"> + <li class="chapter " data-level="6.2.4" data-path="../binaryclass/a9a_minibatch.html"> <a href="../binaryclass/a9a_minibatch.html"> - <b>6.2.3.</b> + <b>6.2.4.</b> - Mini-batch gradient descent + Mini-batch Gradient Descent </a> @@ -1038,7 +1053,7 @@ <b>6.3.1.</b> - Data preparation + Data Preparation </a> @@ -1076,13 +1091,28 @@ </li> - <li class="chapter " data-level="6.3.4" data-path="../binaryclass/news20_adagrad.html"> + <li class="chapter " data-level="6.3.4" data-path="../binaryclass/news20_generic.html"> - <a href="../binaryclass/news20_adagrad.html"> + <a href="../binaryclass/news20_generic.html"> <b>6.3.4.</b> + General Binary Classifier + + </a> + + + + </li> + + <li class="chapter " data-level="6.3.5" data-path="../binaryclass/news20_adagrad.html"> + + <a href="../binaryclass/news20_adagrad.html"> + + + <b>6.3.5.</b> + AdaGradRDA, AdaGrad, AdaDelta </a> @@ -1091,12 +1121,12 @@ </li> - <li class="chapter " data-level="6.3.5" data-path="../binaryclass/news20_rf.html"> + <li class="chapter " data-level="6.3.6" data-path="../binaryclass/news20_rf.html"> <a href="../binaryclass/news20_rf.html"> - <b>6.3.5.</b> + <b>6.3.6.</b> Random Forest @@ -1134,7 +1164,7 @@ <b>6.4.1.</b> - Data preparation + Data Preparation </a> @@ -1185,7 +1215,7 @@ <b>6.5.1.</b> - Data preparation + Data Preparation </a> @@ -1236,7 +1266,7 @@ <b>6.6.1.</b> - Data pareparation + Data Pareparation </a> @@ -1302,7 +1332,7 @@ <b>6.8.1.</b> - Data preparation + Data Preparation </a> @@ -1360,7 +1390,7 @@ <b>7.1.1.</b> - Data preparation + Data Preparation </a> @@ -1375,7 +1405,7 @@ <b>7.1.2.</b> - Data preparation for one-vs-the-rest classifiers + Data Preparation for one-vs-the-rest classifiers </a> @@ -1435,7 +1465,7 @@ <b>7.1.6.</b> - one-vs-the-rest classifier + one-vs-the-rest Classifier </a> @@ -1559,7 +1589,7 @@ <b>8.2.1.</b> - Data preparation + Data Preparation </a> @@ -1567,13 +1597,28 @@ </li> - <li class="chapter " data-level="8.2.2" data-path="../regression/e2006_arow.html"> + <li class="chapter " data-level="8.2.2" data-path="../regression/e2006_generic.html"> - <a href="../regression/e2006_arow.html"> + <a href="../regression/e2006_generic.html"> <b>8.2.2.</b> + General Regessor + + </a> + + + + </li> + + <li class="chapter " data-level="8.2.3" data-path="../regression/e2006_arow.html"> + + <a href="../regression/e2006_arow.html"> + + + <b>8.2.3.</b> + Passive Aggressive, AROW </a> @@ -1610,7 +1655,7 @@ <b>8.3.1.</b> - Data preparation + Data Preparation </a> @@ -1698,7 +1743,7 @@ <b>9.1.1.</b> - Item-based collaborative filtering + Item-based Collaborative Filtering </a> @@ -1734,7 +1779,7 @@ <b>9.2.1.</b> - Data preparation + Data Preparation </a> @@ -1749,7 +1794,7 @@ <b>9.2.2.</b> - LSH/MinHash and Jaccard similarity + LSH/MinHash and Jaccard Similarity </a> @@ -1764,7 +1809,7 @@ <b>9.2.3.</b> - LSH/MinHash and brute-force search + LSH/MinHash and Brute-force Search </a> @@ -1815,7 +1860,7 @@ <b>9.3.1.</b> - Data preparation + Data Preparation </a> @@ -1830,7 +1875,7 @@ <b>9.3.2.</b> - Item-based collaborative filtering + Item-based Collaborative Filtering </a> @@ -1875,7 +1920,7 @@ <b>9.3.5.</b> - SLIM for fast top-k recommendation + SLIM for fast top-k Recommendation </a> @@ -1890,7 +1935,7 @@ <b>9.3.6.</b> - 10-fold cross validation (Matrix Factorization) + 10-fold Cross Validation (Matrix Factorization) </a> @@ -2080,7 +2125,7 @@ <b>13.2.1.</b> - a9a tutorial for DataFrame + a9a Tutorial for DataFrame </a> @@ -2095,7 +2140,7 @@ <b>13.2.2.</b> - a9a tutorial for SQL + a9a Tutorial for SQL </a> @@ -2131,7 +2176,7 @@ <b>13.3.1.</b> - E2006-tfidf regression tutorial for DataFrame + E2006-tfidf Regression Tutorial for DataFrame </a> @@ -2146,7 +2191,7 @@ <b>13.3.2.</b> - E2006-tfidf regression tutorial for SQL + E2006-tfidf Regression Tutorial for SQL </a> @@ -2166,7 +2211,7 @@ <b>13.4.</b> - Generic features + Generic Features </a> @@ -2182,7 +2227,7 @@ <b>13.4.1.</b> - Top-k join processing + Top-k Join Processing </a> @@ -2197,7 +2242,7 @@ <b>13.4.2.</b> - Other utility functions + Other Utility Functions </a> @@ -2352,12 +2397,6 @@ training ; </code></pre> -<p>Hivemall function <a href="../misc/funcs.html#others"><code>hivemall_version()</code></a> shows current Hivemall version, for example:</p> -<pre><code class="lang-sql"><span class="hljs-keyword">select</span> hivemall_version(); -</code></pre> -<blockquote> -<p>"0.5.1-incubating-SNAPSHOT"</p> -</blockquote> <p>Below we list ML and relevant problems that Hivemall can solve:</p> <ul> <li><a href="../binaryclass/general.html">Binary and multi-class classification</a></li> @@ -2603,7 +2642,9 @@ </table> <p>Notice that weight is learned for each possible value in a categorical feature, and for every single quantitative feature.</p> <p>Of course, you can optimize hyper-parameters to build more accurate prediction model. Check the output of the following query to see all available options, including learning rate, number of iterations and regularization parameters, and their default values:</p> -<pre><code class="lang-sql"><span class="hljs-keyword">select</span> train_classifier(<span class="hljs-built_in">array</span>(), <span class="hljs-number">0</span>, <span class="hljs-string">'-help'</span>); +<pre><code class="lang-sql"><span class="hljs-keyword">select</span> train_classifier(<span class="hljs-string">'-help'</span>); +<span class="hljs-comment">-- Hivemall 0.5.2 and before</span> +<span class="hljs-comment">-- select train_classifier(array(), 0, '-help');</span> </code></pre> <h3 id="step-3-prediction">Step 3. Prediction</h3> <p>Now, the table <code>classifier</code> has liner coefficients for given features, and we can predict unforeseen samples by computing a weighted sum of their features.</p> @@ -2625,14 +2666,17 @@ <span class="hljs-comment">-- to join with a model table</span> extract_feature(fv) <span class="hljs-keyword">as</span> feature, extract_weight(fv) <span class="hljs-keyword">as</span> <span class="hljs-keyword">value</span> - <span class="hljs-keyword">from</span> unforeseen_samples t1 LATERAL <span class="hljs-keyword">VIEW</span> explode(features) t2 <span class="hljs-keyword">as</span> fv + <span class="hljs-keyword">from</span> + unforeseen_samples t1 + LATERAL <span class="hljs-keyword">VIEW</span> explode(features) t2 <span class="hljs-keyword">as</span> fv ) <span class="hljs-keyword">select</span> t1.<span class="hljs-keyword">id</span>, sigmoid( <span class="hljs-keyword">sum</span>(p1.weight * t1.<span class="hljs-keyword">value</span>) ) <span class="hljs-keyword">as</span> probability <span class="hljs-keyword">from</span> features_exploded t1 - <span class="hljs-keyword">LEFT</span> <span class="hljs-keyword">OUTER</span> <span class="hljs-keyword">JOIN</span> classifier p1 <span class="hljs-keyword">ON</span> (t1.feature = p1.feature) + <span class="hljs-keyword">LEFT</span> <span class="hljs-keyword">OUTER</span> <span class="hljs-keyword">JOIN</span> classifier p1 + <span class="hljs-keyword">ON</span> (t1.feature = p1.feature) <span class="hljs-keyword">group</span> <span class="hljs-keyword">by</span> t1.<span class="hljs-keyword">id</span> ; @@ -2661,7 +2705,9 @@ <span class="hljs-keyword">id</span>, extract_feature(fv) <span class="hljs-keyword">as</span> feature, extract_weight(fv) <span class="hljs-keyword">as</span> <span class="hljs-keyword">value</span> - <span class="hljs-keyword">from</span> training t1 LATERAL <span class="hljs-keyword">VIEW</span> explode(features) t2 <span class="hljs-keyword">as</span> fv + <span class="hljs-keyword">from</span> + training t1 + LATERAL <span class="hljs-keyword">VIEW</span> explode(features) t2 <span class="hljs-keyword">as</span> fv ), predictions <span class="hljs-keyword">as</span> ( <span class="hljs-keyword">select</span> @@ -2669,7 +2715,8 @@ predictions <span class="hljs-keyword">as</span> ( sigmoid( <span class="hljs-keyword">sum</span>(p1.weight * t1.<span class="hljs-keyword">value</span>) ) <span class="hljs-keyword">as</span> probability <span class="hljs-keyword">from</span> features_exploded t1 - <span class="hljs-keyword">LEFT</span> <span class="hljs-keyword">OUTER</span> <span class="hljs-keyword">JOIN</span> classifier p1 <span class="hljs-keyword">ON</span> (t1.feature = p1.feature) + <span class="hljs-keyword">LEFT</span> <span class="hljs-keyword">OUTER</span> <span class="hljs-keyword">JOIN</span> classifier p1 + <span class="hljs-keyword">ON</span> (t1.feature = p1.feature) <span class="hljs-keyword">group</span> <span class="hljs-keyword">by</span> t1.<span class="hljs-keyword">id</span> ) @@ -2677,10 +2724,13 @@ predictions <span class="hljs-keyword">as</span> ( auc(probability, label) <span class="hljs-keyword">as</span> auc, logloss(probability, label) <span class="hljs-keyword">as</span> logloss <span class="hljs-keyword">from</span> ( - <span class="hljs-keyword">select</span> t1.probability, t2.label - <span class="hljs-keyword">from</span> predictions t1 - <span class="hljs-keyword">join</span> training t2 <span class="hljs-keyword">on</span> (t1.<span class="hljs-keyword">id</span> = t2.<span class="hljs-keyword">id</span>) - <span class="hljs-keyword">ORDER</span> <span class="hljs-keyword">BY</span> probability <span class="hljs-keyword">DESC</span> + <span class="hljs-keyword">select</span> + t1.probability, t2.label + <span class="hljs-keyword">from</span> + predictions t1 + <span class="hljs-keyword">join</span> training t2 <span class="hljs-keyword">on</span> (t1.<span class="hljs-keyword">id</span> = t2.<span class="hljs-keyword">id</span>) + <span class="hljs-keyword">ORDER</span> <span class="hljs-keyword">BY</span> + probability <span class="hljs-keyword">DESC</span> ) t ; </code></pre> @@ -2792,7 +2842,9 @@ predictions <span class="hljs-keyword">as</span> ( </code></pre> <p><code>-loss_function squared</code> means that this query builds a simple linear regressor with the squared loss. Meanwhile, this example optimizes the parameters based on the <code>AdaGrad</code> optimization scheme with <code>l2</code> regularization.</p> <p>Run the function with <code>-help</code> option to list available options:</p> -<pre><code class="lang-sql"><span class="hljs-keyword">select</span> train_regressor(<span class="hljs-built_in">array</span>(), <span class="hljs-number">0</span>, <span class="hljs-string">'-help'</span>); +<pre><code class="lang-sql"><span class="hljs-keyword">select</span> train_regressor(<span class="hljs-string">'-help'</span>); +<span class="hljs-comment">-- Hivemall 0.5.2 and before</span> +<span class="hljs-comment">-- select train_regressor(array(), 0, '-help');</span> </code></pre> <h3 id="step-3-prediction">Step 3. Prediction</h3> <p>Prepare dummy new customers:</p> @@ -2844,7 +2896,9 @@ predictions <span class="hljs-keyword">as</span> ( <span class="hljs-keyword">id</span>, extract_feature(fv) <span class="hljs-keyword">as</span> feature, extract_weight(fv) <span class="hljs-keyword">as</span> <span class="hljs-keyword">value</span> - <span class="hljs-keyword">from</span> training t1 LATERAL <span class="hljs-keyword">VIEW</span> explode(features) t2 <span class="hljs-keyword">as</span> fv + <span class="hljs-keyword">from</span> + training t1 + LATERAL <span class="hljs-keyword">VIEW</span> explode(features) t2 <span class="hljs-keyword">as</span> fv ), predictions <span class="hljs-keyword">as</span> ( <span class="hljs-keyword">select</span> @@ -2943,7 +2997,7 @@ Apache Hivemall is an effort undergoing incubation at The Apache Software Founda <script> var gitbook = gitbook || []; gitbook.push(function() { - gitbook.page.hasChanged({"page":{"title":"Step-by-Step Tutorial on Supervised Learning","level":"5.2","depth":1,"next":{"title":"Binary Classification","level":"6.1","depth":1,"path":"binaryclass/general.md","ref":"binaryclass/general.md","articles":[]},"previous":{"title":"How Prediction Works","level":"5.1","depth":1,"path":"supervised_learning/prediction.md","ref":"supervised_learning/prediction.md","articles":[]},"dir":"ltr"},"config":{"plugins":["theme-api","edit-link","github","splitter","sitemap","etoc","callouts","toggle-chapters","anchorjs","codeblock-filename","expandable-chapters","multipart","codeblock-filename","katex","emphasize","localized-footer"],"styles":{"website":"styles/website.css","pdf":"styles/pdf.css","epub":"styles/epub.css","mobi":"styles/mobi.css","ebook":"styles/ebook.css","print":"styles/print.css"},"pluginsConfig":{"emphasize":{},"callouts":{},"etoc":{"h2lb":3,"header":1,"maxdepth":3,"mindepth":1,"notoc":true},"github":{"url":"https://githu b.com/apache/incubator-hivemall/"},"splitter":{},"search":{},"downloadpdf":{"base":"https://github.com/apache/incubator-hivemall/docs/gitbook","label":"PDF","multilingual":false},"multipart":{},"localized-footer":{"filename":"FOOTER.md","hline":"true"},"lunr":{"maxIndexSize":1000000,"ignoreSpecialCharacters":false},"katex":{},"fontsettings":{"theme":"white","family":"sans","size":2,"font":"sans"},"highlight":{},"codeblock-filename":{},"sitemap":{"hostname":"https://hivemall.incubator.apache.org/"},"theme-api":{"languages":[],"split":false,"theme":"dark"},"sharing":{"facebook":true,"twitter":true,"google":false,"weibo":false,"instapaper":false,"vk":false,"all":["facebook","google","twitter","weibo","instapaper"]},"edit-link":{"label":"Edit","base":"https://github.com/apache/incubator-hivemall/tree/master/docs/gitbook"},"theme-default":{"styles":{"website":"styles/website.css","pdf":"styles/pdf.css","epub":"styles/epub.css","mobi":"styles/mobi.css","ebook":"styles/ebook.css","print":"; styles/print.css"},"showLevel":true},"anchorjs":{"selector":"h1,h2,h3,*:not(.callout) > h4,h5"},"toggle-chapters":{},"expandable-chapters":{}},"theme":"default","pdf":{"pageNumbers":true,"fontSize":12,"fontFamily":"Arial","paperSize":"a4","chapterMark":"pagebreak","pageBreaksBefore":"/","margin":{"right":62,"left":62,"top":56,"bottom":56}},"structure":{"langs":"LANGS.md","readme":"README.md","glossary":"GLOSSARY.md","summary":"SUMMARY.md"},"variables":{},"title":"Hivemall User Manual","links":{"sidebar":{"<i class=\"fa fa-home\"></i> Home":"https://hivemall.incubator.apache.org/"}},"gitbook":"3.x.x","description":"User Manual for Apache Hivemall"},"file":{"path":"supervised_learning/tutorial.md","mtime":"2018-10-18T10:26:56.761Z","type":"markdown"},"gitbook":{"version":"3.2.3","time":"2018-11-13T09:32:29.643Z"},"basePath":"..","book":{"language":""}}); + gitbook.page.hasChanged({"page":{"title":"Step-by-Step Tutorial on Supervised Learning","level":"5.2","depth":1,"next":{"title":"Binary Classification","level":"6.1","depth":1,"path":"binaryclass/general.md","ref":"binaryclass/general.md","articles":[]},"previous":{"title":"How Prediction Works","level":"5.1","depth":1,"path":"supervised_learning/prediction.md","ref":"supervised_learning/prediction.md","articles":[]},"dir":"ltr"},"config":{"plugins":["theme-api","edit-link","github","splitter","sitemap","etoc","callouts","toggle-chapters","anchorjs","codeblock-filename","expandable-chapters","multipart","codeblock-filename","katex","emphasize","localized-footer"],"styles":{"website":"styles/website.css","pdf":"styles/pdf.css","epub":"styles/epub.css","mobi":"styles/mobi.css","ebook":"styles/ebook.css","print":"styles/print.css"},"pluginsConfig":{"emphasize":{},"callouts":{},"etoc":{"h2lb":3,"header":1,"maxdepth":3,"mindepth":1,"notoc":true},"github":{"url":"https://githu b.com/apache/incubator-hivemall/"},"splitter":{},"search":{},"downloadpdf":{"base":"https://github.com/apache/incubator-hivemall/docs/gitbook","label":"PDF","multilingual":false},"multipart":{},"localized-footer":{"filename":"FOOTER.md","hline":"true"},"lunr":{"maxIndexSize":1000000,"ignoreSpecialCharacters":false},"katex":{},"fontsettings":{"theme":"white","family":"sans","size":2,"font":"sans"},"highlight":{},"codeblock-filename":{},"sitemap":{"hostname":"https://hivemall.incubator.apache.org/"},"theme-api":{"languages":[],"split":false,"theme":"dark"},"sharing":{"facebook":true,"twitter":true,"google":false,"weibo":false,"instapaper":false,"vk":false,"all":["facebook","google","twitter","weibo","instapaper"]},"edit-link":{"label":"Edit","base":"https://github.com/apache/incubator-hivemall/tree/master/docs/gitbook"},"theme-default":{"styles":{"website":"styles/website.css","pdf":"styles/pdf.css","epub":"styles/epub.css","mobi":"styles/mobi.css","ebook":"styles/ebook.css","print":"; styles/print.css"},"showLevel":true},"anchorjs":{"selector":"h1,h2,h3,*:not(.callout) > h4,h5"},"toggle-chapters":{},"expandable-chapters":{}},"theme":"default","pdf":{"pageNumbers":true,"fontSize":12,"fontFamily":"Arial","paperSize":"a4","chapterMark":"pagebreak","pageBreaksBefore":"/","margin":{"right":62,"left":62,"top":56,"bottom":56}},"structure":{"langs":"LANGS.md","readme":"README.md","glossary":"GLOSSARY.md","summary":"SUMMARY.md"},"variables":{},"title":"Hivemall User Manual","links":{"sidebar":{"<i class=\"fa fa-home\"></i> Home":"https://hivemall.incubator.apache.org/"}},"gitbook":"3.x.x","description":"User Manual for Apache Hivemall"},"file":{"path":"supervised_learning/tutorial.md","mtime":"2018-12-26T10:16:03.083Z","type":"markdown"},"gitbook":{"version":"3.2.3","time":"2018-12-26T10:20:07.153Z"},"basePath":"..","book":{"language":""}}); }); </script> </div> @@ -2973,7 +3027,7 @@ Apache Hivemall is an effort undergoing incubation at The Apache Software Founda - <script src="https://cdnjs.cloudflare.com/ajax/libs/anchor-js/4.1.1/anchor.min.js";></script> + <script src="../gitbook/gitbook-plugin-anchorjs/anchor.min.js"></script>
