Teach Your Computer to Classify, Train and Make Predictions With Machine Learning

Purpose & objective

Artificial Intelligence is an aspect of computer science that enable computer systems make intelligent decisions like humans. Advances in computing power, has led to generation of enormous amount of data. The challenge becomes what to do with the huge amount of data. Cloud environment where these huge data reside resulted in development of Machine Learning as a tool of Artificial Intelligence that teach computer systems to model, learn, and make predictions from data.
The impact of machine learning is being felt all around us from self-parking cars, driverless cars, intelligent digital assistant, implementing basic adaptive assessment in Google and Microsoft forms mention but a few. Yet, in K-12 school systems, few understand how these mart technologies function; understanding of these novel technologies has largely remained a mirage. The challenge therefore is to introduced AI/ML concepts early in the curriculum at rudimentary or grade appropriate levels to acquaint students with the knowledge, potential and limits that gets them ready for tomorrows jobs that is going to be dominated by AI/ML.

In an anecdotal pre/post survey pilot project conducted with 85 student teachers, the pre-survey indicated they have no idea of what Machine Learning is about. However, after a week (three hours of discussion and demonstrations) they were able to create Machine Learning models of four different facial emotions for a Special Education instruction with 95 – 97% confidence level without out writing a single line of code using no code ML tools such as Teachable Machines and Lobe.
At the end of the proposed workshop/presentation, participants would have classified their data source, train their devices to model data and make predictions. They will go further to integrate the model into a website such as Google site or mobile app to enable users such as students and colleagues test their model.

Outline

During this session which will be interactive and device-based, the following content outline will drive the presentation.
Introduction: (3 minutes)
Participants will discuss current understand of AI/ML to clear misconceptions.
Different Machine Learning tools especially the no-code tools that does not involve computer programming algorithms will be examined
Demonstration (7 minutes)
Participants will be introduced to two Google Teachable Machine and Microsoft Lobe NO-Code Machine Learning web-based and desktop tools, respectively. Participants would have been advised to download Microsoft Lobe(desktop) before the session). They will be required to identify a data source they want to use to train their computer models making predictions. Presenter will demonstrate how to use the tools to train models. Much of resources the participants would need are already provided in the resource link by the presenter before the conference.
Attendee hands-on (30 minutes)
Having identified their data sources, attendees will set about training their models through a simple three step process of 1.) Object classification, 2.) Training and 3.) Prediction. They could use their computer webcam to acquire data or use datasets provided by the presenter.
Conclusion (10 minutes)
Confident in their newly acquired knowledge, attendees will wrap things up by sharing with the audience pedagogic implications of AI/ML in instruction, possibilities

Supporting research

1. Zhang, H., McNeil, S., Gronseth, S., Dogan, B., Handoko, E. & Ugwu, L. (2020). Building Computational Thinking Through Teachable Machine. In D. Schmidt-Crawford (Ed.), Proceedings of Society for Information Technology & Teacher Education International Conference (pp. 127-131). Online: Association for the Advancement of Computing in Education (AACE). Retrieved October 1, 2021 from https://www.learntechlib.org/primary/p/216009/.
2. Heys, J.J. (2018). Machine Learning as a Tool to Identify Critical Assignments. Chemical Engineering Education (CEE), 52(4), 243-250. Retrieved October 1, 2021 from https://www.learntechlib.org/p/189766/.
3. Lane D (2021). Machine Learning for kids: A project-based introduction to artificial intelligence. No Starch Press, CA