Blog AI & ML

Adjusting To New Styles Of Learning – EDUCAUSE 2018

Bob Nilsson Published 2 Nov 2018

An 8-year-old boy craving a cheeseburger hopped into his father’s car and drove himself and his younger sister to a McDonald’s near their Ohio home. The child said he had never driven a vehicle before and quickly learned how by watching YouTube videos. – New York Daily News

Students can learn quickly today, but they learn differently than their parents. New styles of learning were the topic of a growing number of sessions at EDUCAUSE 2018. No less than 22 breakouts dealt with virtual, mixed, or augmented reality.

A group from Case Western Reserve demonstrated live how students in a nearly empty room can collaboratively dissect a human body and practice emergency medical treatment on the victim of a highway accident using mixed reality. The school uses HoloLens headsets, with enhanced software that represents each participant as a full-scale avatar, so the participants know where everyone is. While this real time demo was with a group of five participants, the school has also successfully deployed it with groups as large as 80, all viewing the same virtual objects – in the lecture hall of the future.

[A demo group of five jointly explore a virtual human body during “Applying Mixed Reality to the Classroom of the Future”]

Some of the important benefits of this implementation of mixed reality are no learning time is required to get acquainted with the headset technology, and it can be used with groups that are remotely distributed or local. The project at Case Western Reserve has demonstrated dramatic cost savings (cadavers are expensive) and remarkable improvements in learning and knowledge retention. The shared virtual experience is important for teaching students how to work collaboratively, better preparing them for the actual work environment.

Artificial Intelligence and Machine Learning

Machine learning (ML) takes its cue from how infants learn language; by trying lots of sounds and processing the reinforcing responses from their parents. No baby ever learned language from a text book. Now that sufficient computing power is available, computers can be taught in a similar manner by feeding them massive amounts of experiential data and reinforcing the positive results.

Google brought together a group of schools that have applied ML to university problems with success in the session, “How analytics and Machine Learning are transforming education”. Indiana University fed Canvas class assignment data correlated with student success into ML algorithms and found they could predict student outcomes based on how they were handling their assignments. The school can now automatically reach out to students who are predicted to be in danger and steer them back on track. At Strayer University, the ML system was able to identify students who were having difficulties earlier than the faculty could.

Other areas where AI and ML are being adopted at universities include: predicting which high school students who have contacted the university will actually apply for admission; and providing automated writing feedback, an otherwise labor intensive task. In networking infrastructure, AI and ML are being used to manage and tune Wi-Fi access points, reducing the load on the IT staff.

Related Higher Education Stories