Using Machine Learning for Predicting the Likelihood of Upper Secondary School Student Dropout
According to research there is much room for improving student retention in upper secondary schools in Iceland. Inna is a school administrative and learning management system used by all upper secondary schools in Iceland to some extent. Commissioned in 2001 by the Icelandic Ministry of Education, S...
Main Author: | |
---|---|
Other Authors: | |
Format: | Thesis |
Language: | English |
Published: |
2019
|
Subjects: | |
Online Access: | http://hdl.handle.net/1946/32291 |
id |
ftskemman:oai:skemman.is:1946/32291 |
---|---|
record_format |
openpolar |
spelling |
ftskemman:oai:skemman.is:1946/32291 2023-05-15T16:46:57+02:00 Using Machine Learning for Predicting the Likelihood of Upper Secondary School Student Dropout Hilmar Ævar Hilmarsson 1988- Háskóli Íslands 2019-02 application/pdf image/png http://hdl.handle.net/1946/32291 en eng http://hdl.handle.net/1946/32291 Hugbúnaðarverkfræði Brottfall úr skóla Reiknilíkön Thesis Master's 2019 ftskemman 2022-12-11T06:53:18Z According to research there is much room for improving student retention in upper secondary schools in Iceland. Inna is a school administrative and learning management system used by all upper secondary schools in Iceland to some extent. Commissioned in 2001 by the Icelandic Ministry of Education, Science and Culture, the system has a large digital history of student attendance and school work performance data. The aim of this research is to evaluate the possibility of using machine learning algorithms to predict whether students in upper secondary schools in Iceland are in danger of dropping out of their studies. In this research a feature assessment is done on the data available in Inna. The performance of 13 batch learning supervised machine learning algorithms is evaluated. The algorithms are trained using 28 features from 53674 records of previous student registrations that have been enrolled in matriculation examination studies in Iceland in the years 2003–2018. The best performing classifier after tuning the parameters is Gradient Boosting, followed by Random Forest and AdaBoost. The Gradient Boosting classifier is able to reliably predict with 84% accuracy whether a student is in danger of dropping out or their studies or not. An end user interface for school administrative users to view the classification results for their students currently enrolled in matriculation examination studies is presented. Rannsóknir hafa sýnt fram á að mikið svigrúm er til staðar til að bæta úr brottfalli nemenda úr framhaldsskólum á Íslandi. Inna er stjórnunar- og námskerfi sem notað er af öllum framhaldsskólum á Íslandi á einn veg eða annan. Kerfið var stofnað árið 2001 eftir beiðni frá Mennta- og menningarmálaráðuneytinu og hefur kerfið að geyma langa stafræna sögu um viðveru og frammistöðu nemenda í námi. Tilgangur þessarar rannsóknar er að meta hvort hægt sé að þjálfa reiknigreindarlíkan sem getur spáð fyrir um það hvort nemendur í framhaldsskólum á Íslandi eru í hættu á að detta úr námi. Í rannsókninni eru ... Thesis Iceland Skemman (Iceland) Langa ENVELOPE(-14.220,-14.220,64.626,64.626) |
institution |
Open Polar |
collection |
Skemman (Iceland) |
op_collection_id |
ftskemman |
language |
English |
topic |
Hugbúnaðarverkfræði Brottfall úr skóla Reiknilíkön |
spellingShingle |
Hugbúnaðarverkfræði Brottfall úr skóla Reiknilíkön Hilmar Ævar Hilmarsson 1988- Using Machine Learning for Predicting the Likelihood of Upper Secondary School Student Dropout |
topic_facet |
Hugbúnaðarverkfræði Brottfall úr skóla Reiknilíkön |
description |
According to research there is much room for improving student retention in upper secondary schools in Iceland. Inna is a school administrative and learning management system used by all upper secondary schools in Iceland to some extent. Commissioned in 2001 by the Icelandic Ministry of Education, Science and Culture, the system has a large digital history of student attendance and school work performance data. The aim of this research is to evaluate the possibility of using machine learning algorithms to predict whether students in upper secondary schools in Iceland are in danger of dropping out of their studies. In this research a feature assessment is done on the data available in Inna. The performance of 13 batch learning supervised machine learning algorithms is evaluated. The algorithms are trained using 28 features from 53674 records of previous student registrations that have been enrolled in matriculation examination studies in Iceland in the years 2003–2018. The best performing classifier after tuning the parameters is Gradient Boosting, followed by Random Forest and AdaBoost. The Gradient Boosting classifier is able to reliably predict with 84% accuracy whether a student is in danger of dropping out or their studies or not. An end user interface for school administrative users to view the classification results for their students currently enrolled in matriculation examination studies is presented. Rannsóknir hafa sýnt fram á að mikið svigrúm er til staðar til að bæta úr brottfalli nemenda úr framhaldsskólum á Íslandi. Inna er stjórnunar- og námskerfi sem notað er af öllum framhaldsskólum á Íslandi á einn veg eða annan. Kerfið var stofnað árið 2001 eftir beiðni frá Mennta- og menningarmálaráðuneytinu og hefur kerfið að geyma langa stafræna sögu um viðveru og frammistöðu nemenda í námi. Tilgangur þessarar rannsóknar er að meta hvort hægt sé að þjálfa reiknigreindarlíkan sem getur spáð fyrir um það hvort nemendur í framhaldsskólum á Íslandi eru í hættu á að detta úr námi. Í rannsókninni eru ... |
author2 |
Háskóli Íslands |
format |
Thesis |
author |
Hilmar Ævar Hilmarsson 1988- |
author_facet |
Hilmar Ævar Hilmarsson 1988- |
author_sort |
Hilmar Ævar Hilmarsson 1988- |
title |
Using Machine Learning for Predicting the Likelihood of Upper Secondary School Student Dropout |
title_short |
Using Machine Learning for Predicting the Likelihood of Upper Secondary School Student Dropout |
title_full |
Using Machine Learning for Predicting the Likelihood of Upper Secondary School Student Dropout |
title_fullStr |
Using Machine Learning for Predicting the Likelihood of Upper Secondary School Student Dropout |
title_full_unstemmed |
Using Machine Learning for Predicting the Likelihood of Upper Secondary School Student Dropout |
title_sort |
using machine learning for predicting the likelihood of upper secondary school student dropout |
publishDate |
2019 |
url |
http://hdl.handle.net/1946/32291 |
long_lat |
ENVELOPE(-14.220,-14.220,64.626,64.626) |
geographic |
Langa |
geographic_facet |
Langa |
genre |
Iceland |
genre_facet |
Iceland |
op_relation |
http://hdl.handle.net/1946/32291 |
_version_ |
1766037046072180736 |