The Crime Machine

I’m behind on my podcast listening (well, behind in everything tbh), but Reply All covered the flaws of CompStat, a data system originally employed by the NYPD to track crime and hold officers accountable:

But some of these chiefs started to figure out, wait a minute, the person who’s in charge of actually keeping track of the crime in my neighborhood is me. And so if they couldn’t make crime go down, they just would stop reporting crime. And they found all these different ways to do it. You could refuse to take crime reports from victims, you could write down different things than what had actually happened. You could literally just throw paperwork away. And so that guy would survive that CompStat meeting, he’d get his promotion, and then when the next guy showed up, the number that he had to beat was the number that a cheater had set. And so he had to cheat a little bit more.

I sat in on a CompStat meeting years ago in Los Angeles. I went into it excited to see the data system that helped decrease crime, but I left skeptical after hearing the discussions over such small absolute numbers, which in turn made for a lot of fluctuations percentage-wise. Maybe things are different now a decade later, but I’m not surprised that some intentionally and unintentionally gamed the system.

See also: FiveThirtyEight’s CompStat story from 2015.

Tags: , , ,

Minimizing discrimination in machine learning

From Google Research, a look at how discrimination in machine learning can lead to poor results and what might be done to combat:

Here we discuss “threshold classifiers,” a part of some machine learning systems that is critical to issues of discrimination. A threshold classifier essentially makes a yes/no decision, putting things in one category or another. We look at how these classifiers work, ways they can potentially be unfair, and how you might turn an unfair classifier into a fairer one.

Tags: ,

Discrimination algorithms

Claire Cain Miller for the Upshot on when algorithms discriminate:

There is a widespread belief that software and algorithms that rely on data are objective. But software is not free of human influence. Algorithms are written and maintained by people, and machine learning algorithms adjust what they do based on people’s behavior. As a result, say researchers in computer science, ethics and law, algorithms can reinforce human prejudices.

I bring this up often, because I apparently still hold a grudge, but I will always remember the time I told someone I study statistics. He responded skeptically, "Don't computers do that for you?"

In the words of Jeffrey Heer: "It's an absolute myth that you can send an algorithm over raw data and have insights pop up."

Tags: , ,