Data Mining and Predictive Analysis (eBook)

Intelligence Gathering and Crime Analysis

Colleen McCue (Autor)

eBook Download: EPUB

2006 | 1. Auflage
368 Seiten
Elsevier Science (Verlag)
978-0-08-046462-6 (ISBN)

Infos

It is now possible to predict the future when it comes to crime. In Data Mining and Predictive Analysis, Dr. Colleen McCue describes not only the possibilities for data mining to assist law enforcement professionals, but also provides real-world examples showing how data mining has identified crime trends, anticipated community hot-spots, and refined resource deployment decisions. In this book Dr. McCue describes her use of off the shelf software to graphically depict crime trends and to predict where future crimes are likely to occur. Armed with this data, law enforcement executives can develop risk-based deployment strategies, that allow them to make informed and cost-efficient staffing decisions based on the likelihood of specific criminal activity.

Knowledge of advanced statistics is not a prerequisite for using Data Mining and Predictive Analysis. The book is a starting point for those thinking about using data mining in a law enforcement setting. It provides terminology, concepts, practical application of these concepts, and examples to highlight specific techniques and approaches in crime and intelligence analysis, which law enforcement and intelligence professionals can tailor to their own unique situation and responsibilities.

* Serves as a valuable reference tool for both the student and the law enforcement professional
* Contains practical information used in real-life law enforcement situations
* Approach is very user-friendly, conveying sophisticated analyses in practical terms

---

It is now possible to predict the future when it comes to crime. In Data Mining and Predictive Analysis, Dr. Colleen McCue describes not only the possibilities for data mining to assist law enforcement professionals, but also provides real-world examples showing how data mining has identified crime trends, anticipated community hot-spots, and refined resource deployment decisions. In this book Dr. McCue describes her use of "e;off the shelf"e; software to graphically depict crime trends and to predict where future crimes are likely to occur. Armed with this data, law enforcement executives can develop "e;risk-based deployment strategies,"e; that allow them to make informed and cost-efficient staffing decisions based on the likelihood of specific criminal activity.Knowledge of advanced statistics is not a prerequisite for using Data Mining and Predictive Analysis. The book is a starting point for those thinking about using data mining in a law enforcement setting. It provides terminology, concepts, practical application of these concepts, and examples to highlight specific techniques and approaches in crime and intelligence analysis, which law enforcement and intelligence professionals can tailor to their own unique situation and responsibilities.* Serves as a valuable reference tool for both the student and the law enforcement professional* Contains practical information used in real-life law enforcement situations* Approach is very user-friendly, conveying sophisticated analyses in practical terms

Inhalt

Front Cover 1
Data Mining and Predictive Analysis 5
Copyright Page 6
Contents 9
Foreword 15
Preface 17
Introductory Section 35
1 Basics 37
1.1 Basic Statistics 37
1.2 Inferential versus Descriptive Statistics and Data Mining 38
1.3 Population versus Samples 38
1.4 Modeling 40
1.5 Errors 41
1.6 Overfitting the Model 48
1.7 Generalizability versus Accuracy 48
1.8 Input/Output 51
1.9 Bibliography 52
2 Domain Expertise 53
2.1 Domain Expertise 53
2.2 Domain Expertise for Analysts 54
2.3 Compromise 56
2.4 Analyze Your Own Data 58
2.5 Bibliography 58
3 Data Mining 59
3.1 Discovery and Prediction 61
3.2 Confirmation and Discovery 62
3.3 Surprise 64
3.4 Characterization 65
3.5 “Volume Challenge” 66
3.6 Exploratory Graphics and Data Exploration 67
3.7 Link Analysis 71
3.8 Nonobvious Relationship Analysis (NORA) 71
3.9 Text Mining 73
3.10 Future Trends 74
3.11 Bibliography 74
Methods 77
4 Process Models for Data Mining and Analysis 79
4.1 CIA Intelligence Process 81
4.2 CRISP-DM 83
4.3 Actionable Mining and Predictive Analysis for Public Safety and Security 87
4.4 Bibliography 99
5 Data 101
5.1 Getting Started 103
5.2 Types of Data 103
5.3 Data 104
5.4 Types of Data Resources 105
5.5 Data Challenges 116
5.6 How Do We Overcome These Potential Barriers? 121
5.7 Duplication 122
5.8 Merging Data Resources 123
5.9 Public Health Data 124
5.10 Weather and Crime Data 124
5.11 Bibliography 125
6 Operationally Relevant Preprocessing 127
6.1 Operationally Relevant Recoding 127
6.2 Trinity Sight 128
6.3 Duplication 134
6.4 Data Imputation 134
6.5 Telephone Data 135
6.6 Conference Call Example 137
6.7 Internet Data 144
6.8 Operationally Relevant Variable Selection 145
6.9 Bibliography 148
7 Predictive Analytics 151
7.1 How to Select a Modeling Algorithm, Part I 151
7.2 Generalizability versus Accuracy 152
7.3 Link Analysis 153
7.4 Supervised versus Unsupervised Learning Techniques 153
7.5 Discriminant Analysis 155
7.6 Unsupervised Learning Algorithms 156
7.7 Neural Networks 157
7.8 Kohonan Network Models 159
7.9 How to Select a Modeling Algorithm, Part II 159
7.10 Combining Algorithms 160
7.11 Anomaly Detection 161
7.12 Internal Norms 161
7.13 Defining “Normal” 162
7.14 Deviations from Normal Patterns 164
7.15 Deviations from Normal Behavior 164
7.16 Warning! Screening versus Diagnostic 166
7.17 A Perfect World Scenario 167
7.18 Tools of the Trade 169
7.19 General Considerations and Some Expert Options 171
7.20 Variable Entry 172
7.21 Prior Probabilities 172
7.22 Costs 173
7.23 Bibliography 175
8 Public Safety–Speci.c Evaluation 177
8.1 Outcome Measures 178
8.2 Think Big 183
8.3 Training and Test Samples 187
8.4 Evaluating the Model 192
8.5 Updating or Refreshing the Model 195
8.6 Caveat Emptor 196
8.7 Bibliography 197
9 Operationally Actionable Output 199
9.1 Actionable Output 199
Applications 209
10 Normal Crime 211
10.1 Knowing Normal 212
10.2 “Normal” Criminal Behavior 215
10.3 Get to Know “Normal” Crime Trends and Patterns 216
10.4 Staged Crime 217
10.5 Bibliography 218
11 Behavioral Analysis of Violent Crime 221
11.1 Case-Based Reasoning 227
11.2 Homicide 230
11.3 Strategic Characterization 233
11.4 Automated Motive Determination 237
11.5 Drug-Related Violence 239
11.6 Aggravated Assault 239
11.7 Sexual Assault 240
11.8 Victimology 242
11.9 Moving from Investigation to Prevention 245
11.10 Bibliography 245
12 Risk and Threat Assessment 249
12.1 Risk-Based Deployment 251
12.2 Experts versus Expert Systems 252
12.3 “Normal” Crime 253
12.4 Surveillance Detection 253
12.5 Strategic Characterization 254
12.6 Vulnerable Locations 256
12.7 Schools 257
12.8 Data 261
12.9 Accuracy versus Generalizability 262
12.10 “Cost” Analysis 263
12.11 Evaluation 263
12.12 Output 265
12.13 Novel Approaches to Risk and Threat Assessment 266
12.14 Bibliography 268
Case Examples 271
13 Deployment 273
13.1 Patrol Services 274
13.2 Structuring Patrol Deployment 274
13.3 Data 275
13.4 How To 280
13.5 Tactical Deployment 284
13.6 Risk-Based Deployment Overview 285
13.7 Operationally Actionable Output 286
13.8 Risk-Based Deployment Case Studies 293
13.9 Bibliography 299
14 Surveillance Detection 301
14.1 Surveillance Detection and Other Suspicious Situations 301
14.2 Natural Surveillance 304
14.3 Location, Location, Location 309
14.4 More Complex Surveillance Detection 316
14.5 Internet Surveillance Detection 323
14.6 How To 328
14.7 Summary 330
14.8 Bibliography 331
Advanced Concepts and Future Trends 333
15 Advanced Topics 335
15.1 Intrusion Detection 335
15.2 Identify Theft 336
15.3 Syndromic Surveillance 337
15.4 Data Collection, Fusion and Preprocessing 337
15.5 Text Mining 340
15.6 Fraud Detection 342
15.7 Consensus Opinions 344
15.8 Expert Options 345
15.9 Bibliography 346
16 Future Trends 349
16.1 Text Mining 349
16.2 Fusion Centers 351
16.3 “Functional” Interoperability 352
16.4 “Virtual” Warehouses 352
16.5 Domain-Specific Tools 353
16.6 Closing Thoughts 353
16.7 Bibliography 355
Index 357

Produkt-Details

• Erscheint lt. Verlag: 17.10.2006
• Sprache: englisch
• Themenwelt: Mathematik / Informatik ► Informatik ► Datenbanken
• Themenwelt: Recht / Steuern ► Strafrecht ► Kriminologie
• Themenwelt: Sozialwissenschaften ► Kommunikation / Medien ► Buchhandel / Bibliothekswesen
• Themenwelt: Sozialwissenschaften ► Politik / Verwaltung
• Themenwelt: Sozialwissenschaften ► Soziologie
• Themenwelt: Wirtschaft ► Betriebswirtschaft / Management ► Unternehmensführung / Management
• ISBN-10: 0-08-046462-9 / 0080464629
• ISBN-13: 978-0-08-046462-6 / 9780080464626