50 %
50 %
Information about MarketingPres080313

Published on May 1, 2008

Author: TonyRathburn


Introducing Predictive Analytics:  Introducing Predictive Analytics Business Application of Data Mining Slide2:  It’s a Game… Play to Win! Slide3:  Math as Strategy Slide4:  Past Outcome Past Outcome Past Outcome Model Future Outcomes Why Build Models Slide5:  Understanding the Environment Slide6:  Why Statistics Isn’t Enough Predictive Analytics gives us the opportunity to develop models that more accurately model real data… Slide7:  Traditional Statistics Works… Slide8:  Shifting Our Focus Statistics Concentrates on Central Tendency Gives us an “In General” perspective Required First Step Predictive Analytics Focuses on the Tails of the Distribution Identifies Low Incidence/High Impact Occurrences Slide9:  Predictive Analytics Models Model Any Behavior Best performance in the tails of the distribution One tail… or both… Identify groups that display behaviors at different rates Allocate resources appropriately Classification Use a tool that matches reality Understand what you ‘don’t know’ Slide10:  Predictive Analytics

Add a comment

Related presentations