{"id":646,"date":"2019-10-23T14:25:40","date_gmt":"2019-10-23T01:25:40","guid":{"rendered":"http:\/\/deborahfitchett.com\/blog\/?p=646"},"modified":"2019-10-23T14:25:40","modified_gmt":"2019-10-23T01:25:40","slug":"predicting-student-success-with-leganto-anzreg2019","status":"publish","type":"post","link":"https:\/\/deborahfitchett.com\/blog\/2019\/10\/predicting-student-success-with-leganto-anzreg2019\/","title":{"rendered":"Predicting Student Success with Leganto #anzreg2019"},"content":{"rendered":"<p><strong>Predicting Student Success with Leganto: a \u201cProof of Concept\u201d machine learning project<br \/>\n<\/strong><em>Linda Sheedy, Curtin University<\/em><\/p>\n<p>Early adoptors of Leganto as a reading list solution in 2015 &#8211; mainstreamed in 2017. Now 4700+ reading lists with 115,300 citations viewed 1.5million times by 42,000 students.<\/p>\n<p>Ex Libris proposed a proof of concept project &#8220;to use machine learning to investigate the correlation between student success and activity with the Leganto Reading List&#8221;. Curtin had already been active using learning analytics so thought it would be a good fit.<\/p>\n<p>Business need &#8211; early prediction (within 1-6 weeks) of students who&#8217;ll most likely struggle with their course.<\/p>\n<p>Data:<\/p>\n<ul>\n<li>student profile, grade and academic status data from Curtin &#8211; took significant time and effort to produce this, and inter-department work. Course structure and demographics are complicated.<\/li>\n<li>Leganto usage from Ex Libris<\/li>\n<\/ul>\n<p>Lots of work also combining the datasets.<\/p>\n<p>Function: Ex Libris considered a number of possible algorithms &#8211; currently seems to be settling on the Random Forest algorithm but the final outcome may be a two-stage model.<\/p>\n<p>So far Semester 2 2016 &#8211; Semester 2 2018. So far the algorithm has found the following features are most predictive:<\/p>\n<ul>\n<li>student historical average grades<\/li>\n<li>historical usage engineered feature<\/li>\n<p>weighted student usage per course<\/p>\n<li>student age<\/li>\n<li>student usage in week 1 in relation to class<\/li>\n<\/ul>\n<p>Model total accuracy is 91.9%<br \/>\nRecall: it catches 18.8% of students at risk<br \/>\nPrecision: 69.44% (ie for 10 students predicted at risk, 7 actually will be) &#8211; considered high<\/p>\n<p>The model clearly needs more work &#8211; but increasing recall shouldn&#8217;t be at expense of precision. More data may help along with more tweaking of algorithm.<\/p>\n<p>Project has concluded; not sure where Ex Libris will take the project next or whether it&#8217;ll become a Leganto offering.<\/p>\n<p><strong>Q:<\/strong> What intervention did you take if any?<br \/>\n<strong>A: <\/strong>Just a closed project, all anonymised &#8211; just to see if it&#8217;d work &#8211; so no intervention during this project.<\/p>\n<p><strong>Q:<\/strong> Was demographic data other than age included?<br \/>\n<strong>A:<\/strong> The algorithm found itself that age was a major predictor (other demographic data was included but algorithm didn&#8217;t find it to be predictive of success).<\/p>\n<p><strong>Q:<\/strong> How was analysis improved?<br \/>\n<strong>A: <\/strong>At start of project hoped to prove that students would succeed if they read more. But as it went on it shifted to seeing what predicted when students would struggle.<\/p>\n","protected":false},"excerpt":{"rendered":"<p>Predicting Student Success with Leganto: a \u201cProof of Concept\u201d machine learning project Linda Sheedy, Curtin University Early adoptors of Leganto as a reading list solution in 2015 &#8211; mainstreamed in 2017. Now 4700+ reading lists with 115,300 citations viewed 1.5million times by 42,000 students. Ex Libris proposed a proof of concept project &#8220;to use machine [&hellip;]<\/p>\n","protected":false},"author":1,"featured_media":0,"comment_status":"open","ping_status":"open","sticky":false,"template":"","format":"standard","meta":[],"categories":[1],"tags":[307,298,308],"_links":{"self":[{"href":"https:\/\/deborahfitchett.com\/blog\/wp-json\/wp\/v2\/posts\/646"}],"collection":[{"href":"https:\/\/deborahfitchett.com\/blog\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/deborahfitchett.com\/blog\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/deborahfitchett.com\/blog\/wp-json\/wp\/v2\/users\/1"}],"replies":[{"embeddable":true,"href":"https:\/\/deborahfitchett.com\/blog\/wp-json\/wp\/v2\/comments?post=646"}],"version-history":[{"count":1,"href":"https:\/\/deborahfitchett.com\/blog\/wp-json\/wp\/v2\/posts\/646\/revisions"}],"predecessor-version":[{"id":647,"href":"https:\/\/deborahfitchett.com\/blog\/wp-json\/wp\/v2\/posts\/646\/revisions\/647"}],"wp:attachment":[{"href":"https:\/\/deborahfitchett.com\/blog\/wp-json\/wp\/v2\/media?parent=646"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/deborahfitchett.com\/blog\/wp-json\/wp\/v2\/categories?post=646"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/deborahfitchett.com\/blog\/wp-json\/wp\/v2\/tags?post=646"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}