Predictive Analytics Times
Predictive Analytics Times
Data Reliability and Validity, Redux: Do Your CIO and Data Curators Really Understand the Concepts?
 Here are two recent entries on...
On Variable Importance in Logistic Regression
 The model looks good. It’s parsimonious,...
Data-Driven Decisions for Law Enforcement in Toronto
 For today’s leading deep learning methods...
AI, Machine Learning, and the Basics of Predictive Analytics for Process Management
 APQC Chair Carla O’Dell interviews Predictive...

4 years ago
It is a Mistake to…. Extrapolate

 (Part 8 (of 11) of the Top 10 Data Mining Mistakes, drawn from the Handbook of Statistical Analysis and Data Mining Applications) Modeling “connects the dots” between known cases to build up a plausible estimate of what will happen in related, but unseen, locations in data space. Obviously, models – and especially nonlinear ones — are very unreliable outside the bounds of any known data. (Boundary checks are the very minimum protection against “over-answering”, as discussed in the next installment.) But, there are other types of extrapolations that are equally dangerous. We tend to learn too much from our first

To view this content
Login OR subscribe for free

Already receive the Predictive Analytics Times emails?
As of January 2014, the Predictive Analytics Times now requires legacy email subscribers to upgrade their subscription - one time only - in order to attain a password-protected login and gain complete access.

Click here to complete this one-time subscription upgrade


Existing Users Log In
New User Registration
*Required field

Comments are closed.