Monte F. Hancock, Jr.
Published December 19, 2011
Reference - 302 Pages - 80 B/W Illustrations
ISBN 9781439868362 - CAT# K13109
For Instructors Request Inspection Copy
For Librarians Available on Taylor & Francis eBooks >>
SAVE ~$21.00 on each
Used by corporations, industry, and government to inform and fuel everything from focused advertising to homeland security, data mining can be a very useful tool across a wide range of applications. Unfortunately, most books on the subject are designed for the computer scientist and statistical illuminati and leave the reader largely adrift in technical waters.
Revealing the lessons known to the seasoned expert, yet rarely written down for the uninitiated, Practical Data Mining explains the ins-and-outs of the detection, characterization, and exploitation of actionable patterns in data. This working field manual outlines the what, when, why, and how of data mining and offers an easy-to-follow, six-step spiral process. Catering to IT consultants, professional data analysts, and sophisticated data owners, this systematic, yet informal treatment will help readers answer questions, such as:
Helping you avoid common mistakes, the book describes specific genres of data mining practice. Most chapters contain one or more case studies with detailed projects descriptions, methods used, challenges encountered, and results obtained. The book includes working checklists for each phase of the data mining process. Your passport to successful technical and planning discussions with management, senior scientists, and customers, these checklists lay out the right questions to ask and the right points to make from an insider’s point of view.
Visit the book’s webpage for access to additional resources—including checklists, figures, PowerPoint slides, and a small set of simple prototype data mining tools.
What Is Data Mining and What Can It Do?
A Brief Philosophical Discussion
The Most Important Attribute of the Successful Data Miner: Integrity
What Does Data Mining Do?
What Do We Mean By Data?
The Data Mining Process
Discovery and Exploitation
Eleven Key Principles of Information Driven Data Mining
Key Principles Expanded
Type of Models: Descriptive, Predictive, Forensic
Data Mining Methodologies
A Generic Data Mining Process
RAD Skill Set Designators
Problem Definition (Step 1)
Problem Definition Task 1: Characterize Your Problem
Problem Definition Checklist
Candidate Solution Checklist
Problem Definition Task 2: Characterizing Your Solution
Problem Definition Case Study
Data Evaluation (Step 2)
Data Accessibility Checklist
How Much Data Do You Need?
Methods Used for Data Evaluation
Data Evaluation Case Study: Estimating the Information Content Features
Some Simple Data Evaluation Methods
Data Quality Checklist
Feature Extraction and Enhancement (Step 3)
Introduction: A Quick Tutorial on Feature Space
Characterizing and Resolving Data Problems
Principal Component Analysis
Synthesis of Features
Prototyping Plan and Model Development (Step 4)
Step 4A: Prototyping Plan
Prototyping Plan Case Study
Step 4B: Prototyping/Model Development
Model Development Case Study
Model Evaluation (Step 5)
Evaluation Goals and Methods
What Does Accuracy Mean?
Implementation (Step 6)
Quantifying the Benefits of Data Mining
Tutorial on Ensemble Methods
Getting It Wrong: Mistakes Every Data Miner Has Made
Supervised Learning Genre Section 1—Detecting and Characterizing Known Patterns
Representative Example of Supervised Learning: Building a Classifier
Specific Challenges, Problems, and Pitfalls of Supervised Learning
Recommended Data Mining Architectures for Supervised Learning
Forensic Analysis Genre Section 2—Detecting, Characterizing, and Exploiting Hidden Patterns
Recommended Data Mining Architectures for Unsupervised Learning
Examples and Case Studies for Unsupervised Learning
Tutorial on Neural Networks
Making Syntactic Methods Smarter: The Search Engine Problem
Genre Section 3—Knowledge: Its Acquisition, Representation, and Use
Introduction to Knowledge Engineering
Computing with Knowledge
Inferring Knowledge from Data: Machine Learning
Achieves a unique and delicate balance between depth, breadth, and clarity.
—Stefan Joe-Yen, Cognitive Research Engineer, Northrop Grumman Corporation & Adjunct Professor, Department of Computer Science, Webster University
Used as a primer for the recent graduate or as a refresher for the grizzled veteran, Practical Data Mining is a must-have book for anyone in the field of data mining and analytics.
—Chad Sessions, Program Manager, Advanced Analytics Group (AAG)