Teaching Machine Learning with Applied Interdisciplinary Real World Projects

Gulustan Dogan
Proceedings of the Third Teaching Machine Learning and Artificial Intelligence Workshop, PMLR 207:12-15, 2023.

Abstract

One of the challenges in teaching machine learning to computer science students consists in finding the right hands-on projects that students can work on. Most of the popular Machine Learning projects such as Titanic Survival Prediction and Housing Prices Prediction have many solutions available online. Students working on these problems do not get challenged enough to improve their experimental skills. Moreover, they might follow the methodology in the existing solutions which can discourage them from designing a novel ML solution. Real-world applied projects without online solutions in which students can use their creative problem-solving skills are needed to teach ML courses effectively. Otherwise, students fall into an overfitting problem in which they become ML coders reusing existing codes without ever writing their code. Consequently, we present an approach for creating course project material to achieve this goal. This approach is supported as an applied learning curriculum design methodology with an internal grant at our institution. It is validated experimentally for two semesters in a course taught to both graduate and undergraduate computer science students.

Cite this Paper

BibTeX
@InProceedings{pmlr-v207-dogan23a, title = {Teaching Machine Learning with Applied Interdisciplinary Real World Projects}, author = {Dogan, Gulustan}, booktitle = {Proceedings of the Third Teaching Machine Learning and Artificial Intelligence Workshop}, pages = {12--15}, year = {2023}, editor = {Kinnaird, Katherine M. and Steinbach, Peter and Guhr, Oliver}, volume = {207}, series = {Proceedings of Machine Learning Research}, month = {19--23 Sep}, publisher = {PMLR}, pdf = {https://proceedings.mlr.press/v207/dogan23a/dogan23a.pdf}, url = {https://proceedings.mlr.press/v207/dogan23a.html}, abstract = {One of the challenges in teaching machine learning to computer science students consists in finding the right hands-on projects that students can work on. Most of the popular Machine Learning projects such as Titanic Survival Prediction and Housing Prices Prediction have many solutions available online. Students working on these problems do not get challenged enough to improve their experimental skills. Moreover, they might follow the methodology in the existing solutions which can discourage them from designing a novel ML solution. Real-world applied projects without online solutions in which students can use their creative problem-solving skills are needed to teach ML courses effectively. Otherwise, students fall into an overfitting problem in which they become ML coders reusing existing codes without ever writing their code. Consequently, we present an approach for creating course project material to achieve this goal. This approach is supported as an applied learning curriculum design methodology with an internal grant at our institution. It is validated experimentally for two semesters in a course taught to both graduate and undergraduate computer science students.} }
Endnote
%0 Conference Paper %T Teaching Machine Learning with Applied Interdisciplinary Real World Projects %A Gulustan Dogan %B Proceedings of the Third Teaching Machine Learning and Artificial Intelligence Workshop %C Proceedings of Machine Learning Research %D 2023 %E Katherine M. Kinnaird %E Peter Steinbach %E Oliver Guhr %F pmlr-v207-dogan23a %I PMLR %P 12--15 %U https://proceedings.mlr.press/v207/dogan23a.html %V 207 %X One of the challenges in teaching machine learning to computer science students consists in finding the right hands-on projects that students can work on. Most of the popular Machine Learning projects such as Titanic Survival Prediction and Housing Prices Prediction have many solutions available online. Students working on these problems do not get challenged enough to improve their experimental skills. Moreover, they might follow the methodology in the existing solutions which can discourage them from designing a novel ML solution. Real-world applied projects without online solutions in which students can use their creative problem-solving skills are needed to teach ML courses effectively. Otherwise, students fall into an overfitting problem in which they become ML coders reusing existing codes without ever writing their code. Consequently, we present an approach for creating course project material to achieve this goal. This approach is supported as an applied learning curriculum design methodology with an internal grant at our institution. It is validated experimentally for two semesters in a course taught to both graduate and undergraduate computer science students.
APA
Dogan, G.. (2023). Teaching Machine Learning with Applied Interdisciplinary Real World Projects. Proceedings of the Third Teaching Machine Learning and Artificial Intelligence Workshop, in Proceedings of Machine Learning Research 207:12-15 Available from https://proceedings.mlr.press/v207/dogan23a.html.

Related Material