Comments on “BackerTracker
Chris Gadek@dappermarketer · Head of Growth & Marketing, Doorman
This adds incredible value to the crowd funding space. How accurate has your machine learning algo been this far?
Maxwell Salzberg@maxwellsalz · Co-Founder, BackerKit
@dappermarketer Hi Chris, things are getting tighter as we get more data, but we are finding that once we have 5-6 days of data we are about 80% accurate within a few standard deviations. Right now we are showing an overly conservative spread. What is cool about BackerTracker is that it is unsupervised, meaning that data just needs to be added without any … See more
Maxwell Salzberg@maxwellsalz · Co-Founder, BackerKit
@dappermarketer with that being said what we built certainly is not even close to being perfect, but gives us a great base to build, iterate, and experiment with in the future. Lots of great ideas out there worth playing with!
Chris Gadek@dappermarketer · Head of Growth & Marketing, Doorman
@maxwellsalz that's quite impressive this early on. Well done BackerKit team!