On Saturday, November 19, we hosted our How To Train Your Dragon-themed machine learning workshop at the Mission Valley Library. Led by Angelina Kim, 35 girls explored concepts such as sorting and artificial intelligence. We began by introducing the concept of algorithms, which we illustrated by asking the kids to instruct the volunteers to perform specific tasks (such as sitting or standing) without naming the tasks themselves. One participant commented that this was their favorite part because “it was interesting explaining instructions without being able to say the actual words.” This was followed by a couple of activities about sorting algorithms, specifically bubble sort, and linear and binary search algorithms. We organized pictures of dragons in order from fastest to slowest, watched videos about a Hungarian dance involving bubble sort, and demonstrated linear versus binary searches by organizing participants by height and determining the fastest way to find the girl who was exactly 5 feet tall. The girls learned that binary search is much more efficient than linear search!
After a snack break, we delved into the topic of machine learning itself, discussing artificial intelligence and neural networks. We played tic-tac-toe with a computer, whose playing strategy changed depending on which “player” we selected: they could play offensively, defensively, both, randomly, or use machine learning. The girls learned that the AI that used machine learning or both offensive and defensive playing almost always won, unless it was playing against another player using the same strategy. Finally, we talked about neural networks and used our newfound knowledge to play a modified version of tic-tac-toe. Overall, the girls had a lot of fun, adding that they “will probably go to other workshops in the future.”
0 Comments
|
|