Lesson learned this week: I should really pay more attention to Google Science Fair.
While the rest of the world debated politics and committed crime, a bunch of 13 to 18 year-olds came up with some fantastic submissions for the Fair. The one that caught my eye — naturally — was the winner, Global Neural Network Cloud Service for Breast Cancer.
There are multiple reasons to love this work: it’s a more accurate form of breast cancer detection that works with the raw data provided by biological samples1; it’s up on the cloud so that as much data as possible can be collected; and it’s created by a girl who taught herself programming and neural networks, which she created with Java.
Whew. If I’d done a third of this when I was 17, I’d have been happy with myself.
What’s more, this isn’t the first such competition that Brittany Wenger has participated in. According to the Google slides of her project presentation, she’s been doing similar things for at least the past two years.
But one of the things I find the most arresting is that she put together a neural network simulation from Java, which is the language I program in 90% of the time. I know the basics of how a neural network functions, but to be honest, I hadn’t considered how it would be implemented in object-oriented languages. And this is something Brittany introduces with the deceptively simple “4. Implement a custom neural network in Java.” in her slides. (I’m pretty sure I didn’t know what asynchronous methods were when I was 17. )
I wish I could ask for a peek at her code or something, purely to see how the feedback and weightages would work in a system like this, in Java.
Here’s a nice little interview of her by Scientific American, one of the Fair’s sponsors.
Possibly my favorite part of the interview is this:
I went home that night, and I bought a computer programming book and, with no experience, decided that was what I was going to do with the rest of my life.
This is what I find the most inspiring about the participants of the fair: they find something fascinating, throw themselves at it, and end up scaling up their ideas to solve larger and larger problems.