“It’s near the Free Expression Tunnel.” These words guided me through the expansive NC State campus and muggy ninety-degree heat to room 110 of the Bureau of Mines. Alas, I made it to Laber Labs, and walking through that tunnel represented everything that this program aims to teach me. Laber Labs thrives on the cutting edge intersection of big data and decision making. But its work thrives on the idea of free expression, the notion that individuals deserve to have specialized treatment and outcomes, whatever that looks like, in whatever field.
Of course, language like “treatment” or “outcomes” evokes the clinical setting. And make no mistake. Today, I worked with Eric Rose, a PhD student who specializes in designing optimized patient plans through a series of algorithm-guided choices. Essentially, he seeks to create the ideal system, where given x, and if y, prescribe z. When he succeeds, on average, people will receive the most favorable outcome.
Eric’s work is beautiful and far over my head, replete with eye-scrunching pages of mathematical symbols. He also, though, incorporates these sophisticated methods into computer games that showcase the valuable advances he has made. My biggest achievement of the session was adding a small chunk of code to one such application (visible through clicking on the above link and clicking down on “Games” to the “Flying Squirrel” option):
if (heightKnot1 > heightKnot2){
valley = bisect;
}
else {
valley = bisect + 1;
};
tolerance = 0.1;
if ((noSnakes == 0) & (Math.abs(this.terrain.snakes[nearestSnake].position.x - this.terrain.knotsX[valley]) < tolerance)){
snakeInVal = 1;
}
else {
snakeInVal = 0;
};
features.push(snakeInVal);
Not much, but it’s a start!
