View on Snap!Con
Presented By: Ken Kahn
Abstract:
We’ll demonstrate Snap! blocks for creating, training, and using deep neural networks. The focus will be upon the projects listed here: https://ecraft2learn.github.io/ai/#sample-programs. Examples of the range of projects that can be created with these blocks includes ones that predicts how one might rate the output of a “noisy polygon turtle program”, answers questions about these AI blocks, guesses how confident you are, learns to name colours, predicts the output of mathematical functions, makes scientific predictions, and learns to play games such as Tic Tac Toe.
Here are some slides about this work.
The guide and blocks were designed for high school students though perhaps even younger children can master them. For creating and training models there are both very simple blocks and full-featured ones. See https://ecraft2learn.github.io/ai/ for more information.
