Chapter 12: Using Prepared Data

Chapter 12: Using Prepared Data

Overview

So what?s the benefit of using prepared data to build models? You get more effective models faster. Most of the book so far has described the role of data preparation and how to properly prepare data for modeling. This chapter takes a brief look at the effects and benefits of using prepared data for modeling.

Actually, to examine the preparation results a little more closely, quite often the prepared data models are both better and produced faster. However, sometimes the models are of better quality, but produced in the same time as with unprepared data. Sometimes they are no better but produced a lot faster. For a given data set, both the quality of the model and the speed of modeling depend on the algorithm used and the particular implementation of that algorithm. However, it is almost invariably the case that when data is prepared in the manner described in this book, using the prepared data results in either a better model, a faster model, or a better model faster than when using unprepared data?or than when using data inadequately or improperly prepared. Can this statement be justified?

One credit card company spent more than three months building a predictive model for an acquisition program that generated a response rate of 0.9% over a 2,000,000-piece mailing. Shortly thereafter, using essentially the same modeling tools and data, the company launched another campaign using a model constructed from prepared data. Response rate in this model...

< Previous Excerpt Next Excerpt >

Purchase This Book

Data Preparation for Data Mining

TABLE OF CONTENTS

Chapter 12: Using Prepared Data

Contact Preferences

This is embarrasing...

Customize Your GlobalSpec Experience

Select Your Free Newsletters

Industry Newsletters

Select Your Free Product Alerts

This is embarrasing...