Predicting Customer Pregnancy At Target (How We Would Do It)

blog.custora.com

20 points·pospischil··5 comments

I'd imagine tanimoto or cosine similarity would get you most of the way there while being very off-the-shelf.

If you're going to go the route of binary classification, I'd personally do it via RFs as variable importance (product importance) is built in. (But that's just personal pref)

I think that it's a solid step-by-step thought process on tackling the problem -- I'd probably think of the false positive vs false negative in terms of the relative expected values of success/failure in those classifications. (And perhaps cost-sensitivity could even be added to your original classifier -- perhaps if you had a forest of 500 trees, and you get even 100 votes for pregnant, that's enough to decide to send a pregnancy-targeted mailer)

After watching a coworker innocently ask a woman who wasn't expecting, "When are you due?", I've developed a simple rule for this:

If she tells you she's pregnant: Congratulate her.

If she doesn't: Keep your mouth shut.

Seriously, if you want to target expectant mothers, let them register for a discount program. Diapers are expensive! Any marketing effort that begins, "We think you might be pregnant..." is doomed.

Not much real content here. Feels like a rather crass attempt to capitalize on the buzz around the Target story.

I don't know if crass is the right word. It's opportunistic I guess, but this is how the world of marketing works -- a story about something similar to what you do catches the Internet's brief attention, and you put your hand up and say, hey me too!

Thanks for the feedback, and I'm sorry you found it crass. It is fairly speculative, but we thought it would be interesting to people interested in machine learning and curious about how Target might have approached the problem.

Reply on news.ycombinator.com