Data mining as a torture

Printer-friendly version

In the article Data Mining: torturing the data until they confess Luis Carlos Molina provides a very illuminating on data mining, including examples of interesting applications of it.

Article including the abstract and index, drawn from the same publication:

Summary: The title of this article is an informal explanation of the activity that makes a technology called data mining (data mining). The purpose of this technology is to discover hidden knowledge from large volumes of data. Over the past decade, due to large computational advances, has been incorporated to the organizations to become an essential support when making decisions. Organizations such as corporations, professional sports clubs, universities and governments, among others, make use of this technology as an aid in making their decisions. Some of these examples will be cited in this paper.

CONTENTS

1. Introduction
2. Data mining: concepts and history
3. Applications
    3.1. In government
    3.2. In the
    3.3. In college
    3.4. In space research
    3.5. In sports clubs
4. Extensions of data mining
    4.1. Web mining
    4.2. Text mining
5. Conclusions

To obtain a more graphic content on this, please see the presentation Data mining: Torture the data until they confess, by the same author.

 

 
Google
 
     

Dear, Once again I found a while and decided to put to work on one of the many issues that were pending, it is to continue with the series of video tutorials on DB2. In this case we use IBM Data Studio to create scritps and stored procedures, which will be exposed as Web services on WebSphere Application Server Community Edition (aka WASC). We will do this step by step as usual. Also how to...
This is the presentation used Pere Rovira and Daniel Rodriguez in the Conversion Thursday last Thursday, which was on creating effective Dashboards. Simple, direct and illustrative examples. For those who do not know, on Thursday Conversion is about Web Analytics, while viewing the presentation you'd think is a Business Intelligence event, no?
In the previous installment of this series on how not to build a data warehouse and introduced the concept of "surrogate key." A surrogate key is a unique identifier assigned to each record in a dimension table. This key usually has no specific business sense. They are always numeric.Preferably, an auto-increment integer. Typically, the operating system and uses its own keys, but...