Systems and Methods for User Propensity Classification and Online Auction Design

摘要

Systems, devices, and methods are disclosed for predicting a dynamic floor price for increasing cleared revenue cleared after a winning bid is determined in an online bid auction. The dynamic floor price is predicted from a cascading classifier strategy implemented through a series of cascading machine learning based classifier models that have been trained.