They gather at their East Village temple to listen to their leaders chant arcane codes involving greek letters as they attempt to divine the future.
I am of course referring to the NYC Predictive Analytics monthly meetup held at the 6th floor of 770 Broadway, AOL headquarters.
I stopped by last night to see if I could learn how to create my own digital crystal ball.
I lost my way a few minutes into the introductory lecture on machine learning, given by Carnegie Mellon graduate student Kriti Puniyani.
She delivered what was as described a “gentle introduction” to basic classification techniques. My understanding: if data points are tea leaves then “logistical regression” and “support vectors” are ways to divide the leaves into clumps, with one clump giving an answer of “yes”, and the other “no”.
Predicting successful lectures
So where is all the tea-leaf data coming from?
Many social startups and other companies are swimming in numbers and stats about their particular community’s preferences. The idea is to look for patterns and unleash hidden marketing information.
This task goes to the data whisperers—i.e., math-oriented technicians and computer scientists.
Based on the who’s-hiring segment at the beginning of the meetup, employers could be an energy efficiency startup, a small biomedical imaging firm, a digital ad agency, MTV networks, or NYU’s Biomedical Informatics department.
If Bayesian analysis, maximum likelihood estimators, and predictors are topics you think you’ll need to know to make your company smarter, then the NYC Predictive Analytics meetup is the place to be.
Related articles
- NYC Predictive Analytics Meetup (meetup.com)
- Notes from Tom Mitchell, Chair of CMU’s Machine Learning Department (cmu.edu)
