The slide deck from my 1-hour talk at the Boston FIG Learns conference (Festival of Indie Games). The purpose of the talk was to give a gentle introduction to neural networks, and discuss how they might be used to modify different educational games. 

​

The architecture/framework outlined in the talk is a modified version of adaptNN, which has a link on the project page of this site. 

​

Here is the talk description provided in the schedule: 

"How can you provide personalized user experiences that keep people challenged but not overwhelmed? This talk will discuss how to use modified neural networks to modify game parameters to keep a game challenging but fun!

There will be an intro discussion about how NNs work, how to think about using ML in educational games, and then a brief discussion of AdaptNN — an early-stage implementation of an NN that changes gameplay based on user success in real time.

We’ll conclude with a brief Q&A, and a free form discussion about using ML in educational games."

​

Several attendees afterwards told me how simple my explanations were of different ML concepts, and they felt confident they could go home and begin learning more about artificial intelligence and machine learning. 

​