46,99 €
Mehr erfahren.
- Herausgeber: John Wiley & Sons
- Kategorie: Wissenschaft und neue Technologien
- Serie: Sybex Study Guide
- Sprache: Englisch
Demonstrate your Data Science skills by earning the brand-new CompTIA DataX credential
In CompTIA DataX Study Guide: Exam DY0-001, data scientist and analytics professor, Fred Nwanganga, delivers a practical, hands-on guide to establishing your credentials as a data science practitioner and succeeding on the CompTIA DataX certification exam. In this book, you'll explore all the domains covered by the new credential, which include key concepts in mathematics and statistics; techniques for modeling, analysis and evaluating outcomes; foundations of machine learning; data science operations and processes; and specialized applications of data science.
This up-to-date Study Guide walks you through the new, advanced-level data science certification offered by CompTIA and includes hundreds of practice questions and electronic flashcards that help you to retain and remember the knowledge you need to succeed on the exam and at your next (or current) professional data science role. You'll find:
- Chapter review questions that validate and measure your readiness for the challenging certification exam
- Complimentary access to the intuitive Sybex online learning environment, complete with practice questions and a glossary of frequently used industry terminology
- Material you need to learn and shore up job-critical skills, like data processing and cleaning, machine learning model-selection, and foundational math and modeling concepts
Perfect for aspiring and current data science professionals, CompTIA DataX Study Guide is a must-have resource for anyone preparing for the DataX certification exam (DY0-001) and seeking a better, more reliable, and faster way to succeed on the test.
Sie lesen das E-Book in den Legimi-Apps auf:
Seitenzahl: 616
Veröffentlichungsjahr: 2024
Ähnliche
Table of Contents
Cover
Table of Contents
Title Page
Copyright
Dedication
Acknowledgments
About the Author
About the Technical Editor
Introduction
About the DataX Certification
How This Book Is Organized
Interactive Online Learning Environment and Test Bank
How to Contact the Publisher
Assessment Test
Answers to Assessment Test
Chapter 1: What Is Data Science?
Data Science
Data Science Best Practices
Summary
Exam Essentials
Review Questions
Chapter 2: Mathematics and Statistical Methods
Calculus
Probability Distributions
Inferential Statistics
Linear Algebra
Summary
Exam Essentials
Review Questions
Chapter 3: Data Collection and Storage
Common Data Sources
Data Ingestion
Data Storage
Managing the Data Lifecycle
Summary
Exam Essentials
Review Questions
Chapter 4: Data Exploration and Analysis
Exploratory Data Analysis
Common Data Quality Issues
Summary
Exam Essentials
Review Questions
Chapter 5: Data Processing and Preparation
Data Transformation
Data Enrichment and Augmentation
Data Cleaning
Handling Class Imbalance
Summary
Exam Essentials
Review Questions
Chapter 6: Modeling and Evaluation
Types of Models
Model Design Concepts
Model Evaluation
Summary
Exam Essentials
Review Questions
Chapter 7: Model Validation and Deployment
Model Validation
Communicating Results
Model Deployment
Machine Learning Operations (MLOps)
Summary
Exam Essentials
Review Questions
Chapter 8: Unsupervised Machine Learning
Association Rules
Clustering
Dimensionality Reduction
Recommender Systems
Summary
Exam Essentials
Review Questions
Chapter 9: Supervised Machine Learning
Linear Regression
Logistic Regression
Discriminant Analysis
Naive Bayes
Decision Trees
Ensemble Methods
Summary
Exam Essentials
Review Questions
Chapter 10: Neural Networks and Deep Learning
Artificial Neural Networks
Deep Neural Networks
Summary
Exam Essentials
Review Questions
Chapter 11: Natural Language Processing
Natural Language Processing
Text Preparation
Text Representation
Summary
Exam Essentials
Review Questions
Chapter 12: Specialized Applications of Data Science
Optimization
Computer Vision
Summary
Exam Essentials
Review Questions
Appendix: Answers to Review Questions
Chapter 1: What Is Data Science?
Chapter 2: Mathematics and Statistical Methods
Chapter 3: Data Collection and Storage
Chapter 4: Data Exploration and Analysis
Chapter 5: Data Processing and Preparation
Chapter 6: Modeling and Evaluation
Chapter 7: Model Validation and Deployment
Chapter 8: Unsupervised Machine Learning
Chapter 9: Supervised Machine Learning
Chapter 10: Neural Networks and Deep Learning
Chapter 11: Natural Language Processing
Chapter 12: Specialized Applications of Data Science
Index
End User License Agreement
List of Tables
Chapter 2
TABLE 2.1 Common continuous probability distributions
TABLE 2.2 Common discrete probability distributions
Chapter 3
TABLE 3.1 Common licensing types
Chapter 4
TABLE 4.1 Frequency distribution of grades
TABLE 4.2 Summary of exploratory data analysis methods
Chapter 5
TABLE 5.1 Categorical vehicle color values
TABLE 5.2 One-hot encoded vehicle color values
TABLE 5.3 Ordinal shirt size values
TABLE 5.4 Label encoded shirt size values
TABLE 5.5 Original age values
TABLE 5.6 Age values min-max normalized
TABLE 5.7 Original test scores
TABLE 5.8 Test scores standardized (Z-score)
TABLE 5.9 Exponential population growth data for mice
TABLE 5.10 Log transformed population growth data
TABLE 5.11 Sample age data
TABLE 5.12 Binned sample age data
TABLE 5.13 Monthly sales data by product
TABLE 5.14 Sales data pivoted by month and product
TABLE 5.15 Flattened XML address data
TABLE 5.16 Sample housing data
TABLE 5.17 Sample housing data with engineered variable
Chapter 8
TABLE 8.1 Sample market basket data
Chapter 11
TABLE 11.1 Binary representation of a DTM
TABLE 11.2 Frequency count representation of a DTM
TABLE 11.3 Float-weighted vector representation (TF-IDF) of a DTM
TABLE 11.4 Sample GloVe co-occurrence matrix
Chapter 12
TABLE 12.1 Common applications of computer vision
List of Illustrations
Chapter 1
FIGURE 1.1 Data science, machine learning, and artificial intelligence
FIGURE 1.2 Sales forecast based on historical data
FIGURE 1.3 Using segmentation to identify anomalous data
FIGURE 1.4 Biological network
FIGURE 1.5 Object recognition in computer vision
FIGURE 1.6 The CRISP-DM framework
FIGURE 1.7 The DMBoK framework
FIGURE 1.8 The Jupyter Notebook IDE
Chapter 2
FIGURE 2.1 Curve of showing hypothetical tangent line at
FIGURE 2.2 Area under the curve of for between 0 and 3
FIGURE 2.3 Frequency distribution of the lifespan of sample light bulbs test...
FIGURE 2.4 Probability density function (PDF)
FIGURE 2.5 PDF showing interval of interest (shaded area)
FIGURE 2.6 Cumulative distribution function (CDF)
FIGURE 2.7 Probability mass function (PMF)
FIGURE 2.8 Sampling distributions illustrating the central limit theorem
FIGURE 2.9 A vector in two-dimensional space
FIGURE 2.10 Linearly dependent vectors
FIGURE 2.11 Linearly independent vectors
Chapter 3
FIGURE 3.1 Example of a quantitative survey question
FIGURE 3.2 Relational database schema
FIGURE 3.3 Star schema diagram
FIGURE 3.4 Lottery data in the form of a CSV file
FIGURE 3.5 Lottery data in the form of a TSV file
FIGURE 3.6 Lottery data in the form of a JSON file
FIGURE 3.7 Lottery data in the form of an XML file
FIGURE 3.8 Example of a data lineage diagram
Chapter 4
FIGURE 4.1 Histogram of student math test scores
FIGURE 4.2 Box plot of employee salaries
FIGURE 4.3 Density plot of age distribution
FIGURE 4.4 Quantile-quantile (Q-Q) plot of exam scores against a theoretical...
FIGURE 4.5 Bar chart of the distribution of fruit types
FIGURE 4.6 Bar chart of the average cost per vehicle type
FIGURE 4.7 Scatterplot showing the relationship between salary and years of ...
FIGURE 4.8 Line plot of monthly sales revenue over 12 months
FIGURE 4.9 Sample correlation plot
FIGURE 4.10 Violin plot of the relationship between vehicle type and custome...
FIGURE 4.11 Sankey diagram of sales by region, category, and mode of purchas...
FIGURE 4.12 Cluster visualization of items segmented by average income, popu...
FIGURE 4.13 Sample visualization using principal component analysis (PCA)
FIGURE 4.14 Sample nonstationary monthly sales revenue over a 60-month perio...
FIGURE 4.15 Sample stationary monthly sales revenue over a 60-month period a...
FIGURE 4.16 Sample seasonal monthly sales data over a 60-month period
FIGURE 4.17 Decomposed seasonal monthly sales data showing the trend, season...
FIGURE 4.18 Deseasonalized monthly sales data over a 60-month period
Chapter 5
FIGURE 5.1 Sample skewed distribution before (left) and after (right) being ...
FIGURE 5.2 Union of Table A and Table B
FIGURE 5.3 Intersection of Table A and Table B
FIGURE 5.4 Inner join between Table A and Table B
FIGURE 5.5 Left join between Table A and Table B
FIGURE 5.6 Right join between Table A and Table B
FIGURE 5.7 Full join between Table A and Table B
FIGURE 5.8 Anti-join between Table A and Table B
FIGURE 5.9 Cross join between Table A and Table B
Chapter 6
FIGURE 6.1 Directed acyclic graph showing the relationships between smoking,...
FIGURE 6.2 A sample confusion matrix showing actual versus predicted values...
FIGURE 6.3 The ROC curve for a sample classifier, a perfect classifier, and ...
Chapter 7
FIGURE 7.1 Sample decision tree showing the decision logic for a predictive ...
FIGURE 7.2 Sample feature importance chart for a predictive model
FIGURE 7.3 Sample residual vs. fitted values plot showing linearity
FIGURE 7.4 Sample residual vs. fitted values plot showing heteroscedasticity...
FIGURE 7.5 Sample interactive dashboard
FIGURE 7.6 Sample ML pipeline illustrating Level 0 MLOps maturity
FIGURE 7.7 Sample ML pipeline illustrating Level 1 MLOps maturity
FIGURE 7.8 Sample ML pipeline illustrating Level 2 MLOps maturity
FIGURE 7.9 Model decay monitoring as part of an MLOps pipeline
Chapter 8
FIGURE 8.1 Sample association rule
FIGURE 8.2 k-means clustering result showing five clusters
FIGURE 8.3 The WCSS for clusters with
k
values from 1 to 10
FIGURE 8.4 The average silhouette score for clusters with
k
values from 1 to...
FIGURE 8.5 Dendrogram showing result of hierarchical clustering
FIGURE 8.6 Dendrogram showing the maximum vertical distance between the merg...
FIGURE 8.7 Density-based clustering with DBSCAN
FIGURE 8.8 The curse of dimensionality
FIGURE 8.9 Illustration of a user-item interactions matrix
Chapter 9
FIGURE 9.1 Linear regression line of “best fit”
FIGURE 9.2 Curve of the logistic (sigmoid) function
FIGURE 9.3 Decision boundaries created using LDA (left) and QDA (right) on t...
FIGURE 9.4 Sample decision tree
FIGURE 9.5 Sample decision tree
Chapter 10
FIGURE 10.1 Simple artificial neural network showing the flow of input and o...
FIGURE 10.2 The multilayer perceptron (MLP) showing the input, hidden and ou...
FIGURE 10.3 The threshold activation function
FIGURE 10.4 The sigmoid activation function
FIGURE 10.5 The hyperbolic tangent (tanh) activation function
FIGURE 10.6 The rectified linear unit (ReLU) activation function
Chapter 11
FIGURE 11.1 The continuous bag of words (CBoW) Word2Vec method
FIGURE 11.2 The skip-gram Word2Vec method
Chapter 12
FIGURE 12.1 The feasible region of an optimization problem
FIGURE 12.2 Unconstrained optimization objective function showing potential ...
FIGURE 12.3 Binary image with holes (A) and with the holes filled (B)
FIGURE 12.4 Feature extraction
Guide
Cover
Table of Contents
Title Page
Copyright
Dedication
Acknowledgments
About the Author
About the Technical Editor
Introduction
Begin Reading
Appendix: Answers to Review Questions
Index
End User License Agreement
Pages
v
vi
vii
ix
xi
xiii
xxiii
xxiv
xxv
xxvi
xxvii
xxviii
xxix
xxx
xxxi
xxxii
xxxiii
xxxiv
xxxv
xxxvi
xxxvii
xxxviii
xxxix
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
CompTIA® DataX Study Guide
Exam DY0-001
Fred Nwanganga
Copyright © 2024 by John Wiley & Sons, Inc. All rights, including for text and data mining, AI training, and similar technologies, are reserved.
Published by John Wiley & Sons, Inc., Hoboken, New Jersey.
Published simultaneously in Canada and the United Kingdom.
ISBNs: 9781394238989 (paperback), 9781394239009 (ePDF), 9781394238996 (ePub)
No part of this publication may be reproduced, stored in a retrieval system, or transmitted in any form or by any means, electronic, mechanical, photocopying, recording, scanning, or otherwise, except as permitted under Section 107 or 108 of the 1976 United States Copyright Act, without either the prior written permission of the Publisher, or authorization through payment of the appropriate per-copy fee to the Copyright Clearance Center, Inc., 222 Rosewood Drive, Danvers, MA 01923, (978) 750-8400, fax (978) 750-4470, or on the web at www.copyright.com. Requests to the Publisher for permission should be addressed to the Permissions Department, John Wiley & Sons, Inc., 111 River Street, Hoboken, NJ 07030, (201) 748-6011, fax (201) 748-6008, or online at www.wiley.com/go/permission.
Trademarks: WILEY, the Wiley logo, and Sybex are trademarks or registered trademarks of John Wiley & Sons, Inc. and/or its affiliates, in the United States and other countries, and may not be used without written permission. CompTIA DataX is a trademark of CompTIA, Inc. All other trademarks are the property of their respective owners. John Wiley & Sons, Inc. is not associated with any product or vendor mentioned in this book.
Limit of Liability/Disclaimer of Warranty: While the publisher and author have used their best efforts in preparing this book, they make no representations or warranties with respect to the accuracy or completeness of the contents of this book and specifically disclaim any implied warranties of merchantability or fitness for a particular purpose. No warranty may be created or extended by sales representatives or written sales materials. The advice and strategies contained herein may not be suitable for your situation. You should consult with a professional where appropriate. Further, readers should be aware that websites listed in this work may have changed or disappeared between when this work was written and when it is read. Neither the publisher nor author shall be liable for any loss of profit or any other commercial damages, including but not limited to special, incidental, consequential, or other damages.
For general information on our other products and services, please contact our Customer Care Department within the United States at (800) 762-2974, outside the United States at (317) 572- 3993. For product technical support, you can find answers to frequently asked questions or reach us via live chat at https://sybexsupport.wiley.com.
Wiley also publishes its books in a variety of electronic formats. Some content that appears in print may not be available in electronic formats. For more information about Wiley products, visit our web site at www.wiley.com.
Library of Congress Control Number: 2024940184
Cover image: © Jeremy Woodhouse/Getty Images
Cover design: Wiley
To my darling wife, Melinda, and my A-team (Alex, Abigail, and Andrew).Thank you for your love and support. You make it all worth it!
Acknowledgments
I would like to thank and acknowledge all those who helped directly and indirectly in the development of this book. It takes a lot of hard work and dedication from many people to bring a project like this to completion.
First and foremost, I am profoundly grateful to my family for their unwavering support throughout this demanding project. Your constant encouragement and understanding were crucial as I navigated the complexities of this work. I also wish to express my heartfelt thanks to my friend and colleague, Mike Chapple, who consistently inspires me and encourages me to explore new horizons. A special acknowledgment to Kenyon Brown, the senior acquisitions editor at Wiley. Your guidance and support during this initial collaboration was invaluable. I look forward to many more projects like this one.
To the editing and production team, Brad Jones, Ashirvad Moses, Saravanan Dakshinamurthy, Elizabeth Welch, Arielle Guy, Sara Deichman, and others who worked diligently behind the scenes, thank you for your professionalism, exceptional organizational skills, and the insightful contributions you made toward enhancing the quality of the book. I am also thankful to Dr. Scott Nestler for taking the time to review the content thoroughly and provide detailed, thoughtful technical edits. Your expertise has greatly enhanced the quality of this book, making it a more accurate and valuable resource.
Carole Jelen of Waterside Productions continues to be a great literary agent and partner. Her continued support and ability to develop new opportunities have been tremendously beneficial in bringing this project and others like it to life.
Lastly, to my wonderful student assistants, Melissa Perotin and Ricky Chapple, thank you for reading through the material to make sure that it was coherent and accessible to a broad audience. Your work on the assessment questions was invaluable. I couldn't have done it without you.
About the Author
Fred Nwanganga, PhD, is an author, teacher, and data scientist with more than 20 years of analytics and information technology experience in higher education and the private sector. Fred currently serves as an associate teaching professor in the IT, Analytics, and Operations Department at the University of Notre Dame's Mendoza College of Business. He teaches undergraduate and graduate courses in machine learning, unstructured data analytics, and Python for analytics.
Fred is the author of several LinkedIn Learning courses on machine learning, Python, and generative AI. He is also the coauthor of Practical Machine Learning in R (Wiley, 2020). He earned both his BS and MS in computer science from Andrews University. He also holds an MBA from Indiana University and a PhD in computer science and engineering from the University of Notre Dame.
About the Technical Editor
Scott Nestler is a business analytics “pracademic” (practitioner-academic). Previously, he was director of research & development, as well as principal data scientist and optimization lead, at SumerSports. Prior to that, he was director of statistics & modeling at Accenture Federal Services. Previously, he was the academic director of the MS in Business Analytics program and is still an adjunct associate teaching professor in the Mendoza College of Business at the University of Notre Dame.
Originally from Harrisburg, Pennsylvania, Scott is a 1989 graduate of Lehigh University (with a BS in civil engineering), where he received his commission as an officer through the U.S. Army Reserve Officer Training Corps. He earned a PhD in business and management (management science and finance) from the University of Maryland in 2007 and a Master of Science in applied mathematics and operations research from the Naval Postgraduate School in 1999. He also earned a Master of Strategic Studies from the U.S. Army War College in 2013. He retired from the U.S. Army as a Colonel in 2015. In his last Army assignment, Scott served as director of strategic analytics at the Center for Army Analysis, an internal Army think tank. Scott's other tours of duty include assignments as an assistant professor at the Naval Postgraduate School; director of the center for data analysis and statistics at West Point; chief of strategic assessments at the U.S. Embassy – Baghdad; force structure analyst in the Pentagon; and director of computer operations at West Point. Scott won the Barchi Prize from the Military Operations Research Society in 2010 and was recognized by INFORMS with the Volunteer Award (Gold Level) in 2019. He has earned and maintains the Certified Analytics Professional (CAP) and Accredited Professional Statistician (PStat) certifications. He has published numerous articles and is coauthor (with Wayne Winston and Konstantinos Pelechrinis) of the book Mathletics (Princeton University Press, 2022).
Introduction
Congratulations on taking the initial step toward achieving your CompTIA DataX certification. The DataX certification, as described by CompTIA, is “the premier skills development program for highly experienced professionals seeking to validate their competency in the rapidly evolving field of data science.” This study guide is tailored for data scientists who are in the early to mid-stages of their careers. It is designed to serve as a refresher for some and a source of new insights for others. No matter your level of expertise, this guide aims to solidify your understanding of essential data science tools and concepts necessary to effectively prepare for and pass the DataX certification exam.
In the following pages, you will find essential information about the CompTIA DataX exam, details on the organization and scope of this book, and a sample assessment test. This test is intended to help gauge your initial readiness for the certification exam. The answer key for the assessment questions references which chapter within the book addresses the concepts or exam objective behind the question. I encourage you to concentrate your study efforts on those chapters that cover areas where you feel you need to build your skills and confidence.
About the DataX Certification
The DataX certification is designed to be a vendor-neutral validation of expert-level data science skills. CompTIA recommends the certification for professionals with 5+ years of experience in data science or similar roles. You can find additional information about the certification at:
www.comptia.org/certifications/datax
According to CompTIA, the certification is designed to assess a candidate's ability to:
Understand and implement data science operations and processes
Apply mathematical and statistical methods appropriately and understand the importance of data processing and cleaning, statistical modeling, linear algebra, and calculus concepts
Apply machine learning models and understand deep learning concepts
Utilize appropriate analysis and modeling methods and make justified model recommendations
Demonstrate understanding of industry trends and specialized data science applications
Certification Objectives
CompTIA goes to great lengths to ensure that its certifications accurately reflect industry best practices. It works with a team of professionals, training providers, publishers, and subject matter experts (SMEs) to establish baseline competency for each of its exams. Based on this information, CompTIA has published five major domains that the DataX certification exam covers. The following is a list of the domains and the extent to which they are represented on the certification exam:
Domain
Percentage of exam
1.0 Mathematics and Statistics
17%
2.0 Modeling, Analysis, and Outcomes
24%
3.0 Machine Learning
24%
4.0 Operations and Processes
22%
5.0 Specialized Applications of Data Science
13%
Certification Exam
The DataX exam employs what CompTIA refers to as a “performance-based assessment” format. This approach integrates traditional multiple-choice questions with a variety of interactive question types, including fill-in-the-blank, multiple-response, drag-and-drop, and image-based problems, to create a more dynamic and comprehensive evaluation of a candidate's abilities. For more details about CompTIA's performance exams, visit:
www.comptia.org/testing/testing-options/about-comptia-performance-exams
The exam consists of 90 questions and has a time limit of 165 minutes. The results are provided in a pass/fail format. As you prepare, keep in mind two important aspects regarding the nature of the questions you will encounter.
First, CompTIA exams are known for their occasionally ambiguous questions. You may find yourself faced with multiple answers that seem correct, requiring you to choose the “most correct” one based on your knowledge and sometimes intuition. It's important not to spend too much time on these questions. Make your best choice, and then move on to the next question.
Second, be aware that CompTIA often includes unscored questions in their exams to collect psychometric data, a process known as item seeding. These questions are used to help develop future versions of the exam. Although these questions won't affect your score, you may not be able to distinguish them from scored questions, so you should attempt to answer every question as accurately as possible. Before starting the exam, you'll be informed about the possibility of encountering unscored questions. If you come across a question that doesn't seem related to any of the stated exam objectives, it might be one of these seeded questions, but since you can't be sure, it's best to treat every question as if it counts toward your final score.
Taking the Exam
Once you are ready to take the exam, visit the CompTIA store (https://store.comptia.org) to purchase a voucher for the exam. This book also includes a coupon that you may use to save 10 percent on the exam registration. CompTIA offers two options for taking the certification exam. You can either take the exam in person at a Pearson VUE testing center or online. The online exam involves a remote exam proctoring service powered by Pearson OnVUE.
You can find more information about CompTIA testing options at www.comptia.org/testing/testing-options/about-testing-options.
How This Book Is Organized
This study guide covers everything you need to prepare and pass the DataX exam. Each chapter includes several recurring elements to help you prepare. Here's a description of some of those elements:
Assessment Test
At the conclusion of this introduction, you'll find an assessment test designed to gauge your readiness for the exam. I recommend taking this test before you begin reading the book, as it will help you identify which areas might require further review. The answers to the assessment test questions are provided at the end of the test. Each answer comes with an explanation and a note indicating the chapter where the relevant material is covered, allowing you to focus your studies more effectively.
Summary
The summary at the end of each chapter provides a concise review, highlighting the key points and concepts discussed. This overview helps to reinforce your understanding and ensures you grasp the essential elements covered in the chapter.
Exam Essentials
The “Exam Essentials” section located near the end of each chapter underscores topics that are likely to be included on the exam in some capacity. While it's impossible to predict the exact content of the certification exam, this section emphasizes crucial concepts that are fundamental to understanding the topics discussed in the chapter. This feature is designed to reinforce your knowledge and help you focus on the most significant aspects that could be tested.
Chapter Review Questions
Each chapter includes 20 practice questions intended to assess your understanding of the key ideas discussed. After completing each chapter, take the time to answer these questions. If you find some of your responses are incorrect, it's a signal that you should revisit and spend additional time on those topics. The answers to the practice questions are located in Appendix. Please note that these questions are designed to measure your retention of the material and may not necessarily mirror the format or complexity of the questions you will encounter on the exam.
The chapters in this book are structured to facilitate a smooth flow and deepen your understanding of key concepts. They are not necessarily arranged in alignment with the sequence or structure of the certification exam objectives. To assist you in your exam preparation, the following is a high-level map that shows how the exam objectives correspond to the chapters in this study guide. This mapping will help you navigate the material more effectively and ensure that you cover all necessary topics as you prepare for the exam.
Exam objective
Chapter(s)
1.0 Mathematics and Statistics
1.1 Given a scenario, apply the appropriate statistical method or concept.
2
,
6
1.2 Explain probability and synthetic modeling concepts and their uses.
2
1.3 Explain the importance of linear algebra and basic calculus concepts.
2
1.4 Compare and contrast various types of temporal models.
6
2.0 Modeling, Analysis, and Outcomes
2.1 Given a scenario, use the appropriate exploratory data analysis (EDA) method or process.
4
2.2 Given a scenario, analyze common issues with data.
4
2.3 Given a scenario, apply data enrichment and augmentation techniques.
5
2.4 Given a scenario, conduct a model design iteration process.
7
2.5 Given a scenario, analyze results of experiments and testing to justify final model recommendations and selection.
7
2.6 Given a scenario, translate results and communicate via appropriate methods and mediums.
7
3.0 Machine Learning
3.1 Given a scenario, apply foundational machine learning concepts.
6
,
8
,
9
,
10
3.2 Given a scenario, apply appropriate statistical supervised machine learning concepts.
9
3.3 Given a scenario, apply tree-based supervised machine learning concepts.
9
3.4 Explain concepts related to deep learning.
10
3.5 Explain concepts related to unsupervised machine learning.
8
4.0 Operations and Processes
4.1 Explain the role of data science in various business functions.
1
4.2 Explain the process of and purpose for obtaining different types of data.
3
4.3 Explain data ingestion and storage concepts.
3
4.4 Given a scenario, implement common data-wrangling techniques.
5
4.5 Given a scenario, implement best practices throughout the data science life cycle.
1
4.6 Explain the importance of DevOps and MLOps principles in data science.
7
4.7 Compare and contrast various deployment environments.
7
5.0 Specialized Applications of Data Science
5.1 Compare and contrast optimization concepts.
12
5.2 Explain the use and importance of natural language processing (NLP) concepts.
11
5.3 Explain the use and importance of computer vision concepts.
12
5.4 Explain the purpose of other specialized applications in data science.
1
Exam objectives are subject to change by CompTIA at any time without prior notice. Always endeavor to visit the CompTIA website (www.comptia.org) for the most current exam objectives.
Interactive Online Learning Environment and Test Bank
This book comes with a number of interactive online learning tools to help you prepare for the certification exam. Here's a description of some of those tools:
Bonus Practice Exams
In addition to the practice questions provided for each chapter, this study guide features two practice exams. These exams are designed to test your knowledge of the material covered throughout the book, allowing you to assess your readiness for the actual exam and identify areas where you may need further study.
Sybex Test Preparation Software
Sybex's test preparation software enhances your study experience by offering electronic versions of the review questions from each chapter, along with bonus practice exams. With this software, you can customize your preparation by building and taking tests that focus on specific domains, individual chapters, or the entire range of DataX exam objectives through randomized tests. This flexibility allows you to tailor your study approach to best suit your needs and ensure comprehensive coverage of the material.
Electronic Flashcards
This study guide includes over 100 flashcards designed to reinforce your learning and facilitate last-minute test preparation before the exam. These flashcards are a valuable tool for reviewing key concepts and ensuring you are well prepared for testing day.
Go to www.wiley.com/go/sybextestprep to register and gain access to this interactive online learning environment and test bank with study tools.
Like all exams, the DataX certification from CompTIA is updated periodically and may eventually be retired or replaced. At some point after CompTIA is no longer offering this exam, the old editions of our books and online tools will be retired. If you have purchased this book after the exam was retired, or are attempting to register in the Sybex online learning environment after the exam was retired, please know that we make no guarantees that this exam's online Sybex tools will be available once the exam is no longer available.
How to Contact the Publisher
If you believe you have found a mistake in this book, please bring it to our attention. At John Wiley & Sons, we understand how important it is to provide our customers with accurate content, but even with our best efforts an error may occur.
In order to submit your possible errata, please email it to our Customer Service Team at [email protected] with the subject line “Possible Book Errata Submission.”
Assessment Test
A technology firm is developing a new app that uses biometric data. To prevent the misuse of this sensitive information, which of these techniques should be prioritized to secure the data?
Increasing server capacity for data storage
Making sure users have strong passwords
Implementing robust data anonymization processes
Enhancing user interface security features
Ebube is analyzing a company's logistics operations to improve delivery times. In which step in the requirements-gathering process would she identify key metrics like average delivery time and percentage of on-time deliveries?
Defining business objectives
Understanding business processes
Determining the project's budget
Conducting cost-benefit analyses
A cybersecurity firm wants to detect unusual network traffic that could indicate a security breach. Which of these applications of data science is best suited for this?
Natural language processing
Recommendation systems
Prediction
Segmentation
Yucheng is conducting a study to analyze the distribution of wealth among individuals in a country. The distribution is expected to have a few individuals with extremely high wealth compared to the majority. Which of these probability distributions is most appropriate for modeling the data?
Continuous uniform
Student's t
Power law
Gaussian
What is a two-sample t-test used for?
To compare the means of two independent groups to determine if there is a significant difference
To compare the mean of a single sample to a known population mean
To compare the means of two related groups or samples at two points in time
To compare the means of more than two independent groups
Verite is examining a distribution of stock returns. The distribution has a longer tail on the left side compared to the right side. How should he characterize this distribution in terms of skewness?
Positively skewed
Negatively skewed
Zero skewness
Right skewed
Migdalia wants to estimate how much time customers spend on average shopping in a chain of retail stores. To do this, she tracks the shopping time for a sample of 500 customers and calculates the average. In this scenario, the average shopping time calculated from the sample is an example of:
A parameter
A hypothesis
A confidence interval
A statistic
Kevin wants to detect lightning strikes as soon as they occur using an array of sensors spread across a 25-mile radius from his base station. Which data ingestion approach should he use and why?
Batching, because it is cost effective
Batching, because the data can be ingested after a predetermined time interval has elapsed
Streaming, because he can receive real-time alerts
Streaming, because the data can be aggregated before storage
Which of the following datasets would be the most suitable candidate for compression to improve storage efficiency without significantly impacting data retrieval performance?
Real-time telemetry data from an autonomous vehicle
Daily atmospheric pressure readings from a weather station
Instantaneous stock trade data for high-frequency trading algorithms
Live video feed from a security camera
Which of the following formats is specifically designed for organizing and storing large quantities of structured scientific data?
JSON
XML
YAML
HD5
Which of the following is not an appropriate way to handle missing data?
Remove the missing records.
Replace the missing data with the mean of the non-missing values of the same feature.
Use machine learning to predict the value of the missing data.
Replace the missing data with random values.
Pete maintains a baseball database containing information and statistics on every player from the last decade. One column of Pete's database is the player's team. Which type of variable is this?
Continuous
Discrete
Nominal
Ordinal
Professor Held teaches a college course with over 300 students. He has two separate lists in his possession. One list is of students who received an A on the midterm exam, and the other is a list of students who received an A on the final exam. Which type of join should Professor Held use to create a list of students who received an A on both exams?
A left join
An inner join
An anti-join
A cross join
Which of the following techniques results in values with a mean of 0 and a standard deviation of 1?
Log transformation
Box-Cox transformation
Binning
Standardization
Sally converts nested data in JSON format to tabular form so she can more easily work with it. Which of the following does she do?
Pivoting
Flattening
Ground truth labeling
Binning
Naliba works for a travel agency and would like to predict how many flights are likely to be canceled each day over the next six months. She has access to the daily flight cancelation data for the past five years. Which of these models would be most appropriate to make this forecast?
Linear regression
Binary classification
ARIMA
Survival analysis
Ahmed has created a model to predict how many games a football team is likely to win in the coming season. The model performs very well on the training data but does poorly on the test data. Which of the following should Ahmed consider doing to remedy this?
Introduce cross-validation to the model training process.
Reduce the number of predictors in the model.
Add more predictors to the model.
Tune the model hyperparameters.
A hospital is developing a model to help classify tumors as either malignant (cancerous) or benign. Assuming that malignant is the class of interest in this model, which of these metrics should the model prioritize for maximization?
Sensitivity
Specificity
Area under the curve (AUC)
Accuracy
A healthcare organization has developed a machine learning model to predict the risk of readmission based on patient characteristics. They want to share the model's insights with a group of clinicians who are not familiar with machine learning concepts. Which of these visualization tools would be most appropriate?
An interactive dashboard
A decision tree visualization
A confusion matrix
A feature importance chart
In an MLOps workflow, which of the following best describes the purpose of continuous monitoring?
To automate the deployment of new models to production
To regularly update the model with new data to maintain its performance
To streamline the data preprocessing and feature engineering stages
To ensure the security and compliance of the deployed models
Which of the following represents a challenge associated with hybrid deployment?
Sensitive data cannot be retained on premises.
Scalability offered by the cloud cannot be leveraged.
Ensuring seamless integration between cloud and on-premises environments can be complex.
Hybrid deployment requires a greater investment in physical infrastructure compared to other deployment methods.
Sangita works for an online video streaming startup. She wants to create an algorithm that will recommend new videos to users based on their past viewing history. Which of the following techniques is best for this task?
Association rules
Clustering analysis
Dimensionality reduction
Content-based filtering
Which of the following is not a typical reason to conduct principal component analysis (PCA)?
To minimize the dimensionality of a dataset
To improve the interpretability of a model
To minimize the risk of overfitting
To improve the efficiency of a model
A grocery store is analyzing historical customer purchases to identify which items are frequently bought together. Based on their analysis, they find that customers who buy both cheese and bread are more likely to also buy lunch meat. Which type of unsupervised machine learning approach are they using?
Association rules
Recommender systems
Clustering
Dimensionality reduction
Tori plans to use linear regression to predict car prices. She applies the Durbin–Watson test to all the observations in her historical dataset. Which linear regression assumption is she trying to validate?
Autocorrelation of residuals
Homoscedasticity
Independence of observations
Normality of residuals
Fatima wants to create a linear regression model to predict the grades of students in a college course. However, she has too many predictors and wants to reduce them. Which of these techniques should she use?
L2 regularization
Ridge regularization
L1 regularization
Gradient descent
Sanjay wants to predict the outcome of basketball games. He builds an ensemble model that combines the results of a logistic regression model and a decision tree to make predictions. Which approach is he using?
Bagging
Stacking
Boosting
Bootstrap aggregating
DJ is a marketing analyst for a grocery store chain and would like to categorize shoppers into three categories: loyal customers, occasional buyers, and one-time customers. He is using a neural network for this classification problem. Which activation function should he use in the output layer?
Threshold
SoftMax
Sigmoid
Hyperbolic tangent
Which of these approaches should Grace use to prevent her neural network model from overfitting against the training data?
Batch normalization
Learning rate schedulers
Early stopping
Vanishing gradients
Vamsi is in the process of creating a large language model to enhance the customer service chatbot on his company's website. Which deep learning architecture is most suitable for this purpose?
Generative adversarial network
Convolutional neural network
Transformer
Recurrent neural network
Joy is exploring a large collection of news articles to discover the underlying thematic structure. She wants to identify sets of words that frequently occur together and assign each article to one or more of these sets. Which text analysis technique is Joy using?
Keyword extraction
Sentiment analysis
Topic modeling
Semantic matching
Alex wants to automatically create product descriptions for an online product catalog based on specific inputs such as product features and specifications. Which of these aspects of natural language processing is most relevant to Alex's goal?
Language understanding
Language generation
Named entity recognition
Semantic analysis
Patrick is developing a search engine that retrieves documents that are contextually related to a user's query, even if the exact query terms are not present in the document. Which of these is most relevant to Patrick's task?
Semantic matching
Sentiment analysis
Topic modeling
String matching
The one-armed bandit problem is often used as a simplified model for decision-making in various fields. In which of the following scenarios can the one-armed bandit problem be applied as a model for optimization?
Determining the optimal mix of crops to plant on a farm
Allocating budget among different marketing channels
Scheduling flights to minimize delays
Selecting the best treatment option for a patient
A security system needs to use facial recognition to verify the identity of individuals entering a building. Which computer vision approach is primarily involved in this application?
Object detection and recognition
Image segmentation
Optical character recognition (OCR)
Motion analysis and object tracking
A farm wants to optimize its irrigation water usage to maximize crop yield while adhering to regulations. What type of optimization problem is this?
Pricing
Network topology
Scheduling
Resource allocation
Answers to Assessment Test
C. For an app using sensitive biometric data, implementing robust data anonymization processes is essential to secure the data against misuse and ensure privacy and compliance. See
Chapter 1
for more information.
B. Ebube would identify critical metrics such as average delivery time and percentage of on-time deliveries, which are pivotal for analyzing and improving logistics operations, during the “understanding business processes” phase. See
Chapter 1
for more information.
D. Anomaly detection based on segmentation is particularly effective in identifying unusual data points or patterns, such as those that might indicate a cybersecurity threat. This technique can help the company quickly isolate and respond to a potential security breach. See
Chapter 1
for more information.
C. The power law distribution is characterized by a heavy tail and is used when one quantity varies as a power of another. It is suitable for modeling the distribution of wealth, where a few individuals have significantly higher wealth than the majority. See
Chapter 2
for more information.
A. A two-sample t-test, also known as the independent samples t-test, is used to determine whether there is a significant difference between the means of two independent groups. See
Chapter 2
for more information.
B. The distribution would be characterized as negatively skewed because the left tail is longer or heavier than the right. In a negatively skewed distribution, the majority of the data is concentrated on the right side, with a few extreme values on the left. See
Chapter 2
for more information.
D. The average shopping time calculated from the sample is a statistic, as it is a numerical characteristic of the sample used to estimate the corresponding population parameter. See
Chapter 2
for more information.
C. Streaming is the most appropriate method of ingestion in this scenario. Streaming would enable Kevin to capture and analyze each sensor's data instantaneously, providing the ability to react to lightning strikes as they happen. See
Chapter 3
for more information.
B. Compression introduces a delay in the data access pipeline. Daily atmospheric pressure readings from a weather station, while valuable, do not typically require the instant access that real-time systems, like the other options in the question, demand. See
Chapter 3
for more information.
D. Unlike JSON, XML, and YAML, which are more suited for semi-structured data, HDF5 is a binary file format that provides a versatile and efficient methodology for organizing and storing complex scientific datasets that demand a structured storage approach. See
Chapter 3
for more information.
D. Substituting missing data with random values is not an advisable strategy, as it can inject random noise into the dataset, potentially skewing analysis and outcomes. See
Chapter 4
for more information.
C. Variables can be broken down into two categories, quantitative and qualitative. Team name is a qualitative variable because it is not numerical. Qualitative variables can either be nominal or ordinal. Because there is no inherent order among team names, it is considered a nominal variable. See
Chapter 4
for more information.
B. An inner join will merge the two lists based on common entries, thus displaying only those students who earned an A grade on both the midterm and the final exams. See
Chapter 5
for more information.
D. Standardization, often referred to as Z-score normalization, is a scaling technique that transforms features to have a mean of 0 and a standard deviation of 1. See
Chapter 5
for more information.
B. Flattening refers to the process of transforming hierarchical or multilevel structured data into a flat, tabular format. See
Chapter 5
for more information.
C. Because Naliba is working with chronological data, she should create a time-series model. ARIMA, short for autoregressive integrated moving average, is a time-series model that factors in historical values and forecast errors. See
Chapter 6
for more information.
B. Ahmed's model appears to be overfitting the training data. It is not generalizing well to the test data. Using feature selection to reduce the number of predictors in the model is one way to address this. See
Chapter 6
for more information.
A. Sensitivity measures the ability of the model to correctly identify malignant tumors. High sensitivity means that the model is effective at catching malignant cases, which is crucial in a medical context where early detection of cancer can significantly impact treatment success and patient survival. See
Chapter 6
for more information.
A. An interactive dashboard with drill-down capabilities would allow clinicians to explore not only the overall predictions of the model but also the specific relationships between multiple features and readmission risk. See
Chapter 7
for more information.
B. Continuous monitoring and model retraining are crucial in an MLOps workflow to keep the model updated with fresh data and maintain its accuracy and relevance over time. This process helps address concept drift and data drift, ensuring the model continues to perform well on new data. See
Chapter 7
for more information.
C. One of the primary challenges of hybrid deployment is achieving seamless integration between cloud and on-premises environments. This involves ensuring consistent data management, security protocols, and application performance across both platforms, which can be complex and requires careful planning and coordination. See
Chapter 7
for more information.
D. Content-based filtering is the most appropriate technique for this task, as it uses the characteristics of items (in this case, videos) that users have previously interacted with to recommend similar items. See
Chapter 8
for more information.
B. PCA is commonly used to reduce the dimensionality of a dataset, decrease the risk of overfitting, and enhance the efficiency of a model. However, improving the interpretability of a model is not a primary reason for conducting PCA, as the transformation to principal components can sometimes make the data more abstract and less directly interpretable in terms of the original features. See
Chapter 8
for more information.
A. This is an example of the use of association rules, an unsupervised machine learning approach that describes the co-occurrence of items within a transaction set. See
Chapter 8
for more information.
A. Tori is validating the independence of residuals assumption of linear regression, which states that the residuals from the regression should not be correlated with each other. The Durbin–Watson test is primarily used to detect the presence of autocorrelation among the residuals, a specific form of independence check. See
Chapter 9
for more information.
C. L1 regularization (LASSO regression) modifies the loss function to include a penalty that can reduce some coefficients to zero, effectively removing them from the model. Therefore, it should be used if feature selection is a priority for Fatima. See
Chapter 9
for more information.
B. Stacking involves combining the predictions of multiple heterogenous base models using a meta-model. In Sanjay's case, the logistic regression model and the decision tree serve as the base models, and their predictions are combined to make the final prediction. See
Chapter 9
for more information.
B. The SoftMax activation function is particularly useful for multiclass classification. The function returns a decimal probability for each class, allowing the model to assign each item to its most probable class. See
Chapter 10
for more information.
C. Early stopping is a regularization technique used to prevent overfitting in neural networks. It involves monitoring the model's performance on a validation set and stopping the training process when the performance starts to degrade or no longer improves significantly. This prevents the model from learning the noise in the training data, which is a common cause of overfitting. See
Chapter 10
for more information.
C. The Transformer architecture is particularly well suited for building large language models used in natural language processing tasks, including chatbots. It excels at handling sequential data, such as text, and can process entire sentences or even paragraphs in parallel, significantly improving efficiency and effectiveness over traditional models. See
Chapter 10
for more information.
C. Topic modeling is an unsupervised machine learning technique used to discover the underlying thematic structure in a large collection of documents by identifying topics (sets of words that frequently occur together) and assigning each document to one or more of these topics. See
Chapter 11
for more information.
B. Language generation can be used in automated content creation to create written content for websites, reports, and articles based on specific inputs or prompts. See
Chapter 11
for more information.
A. Semantic matching involves comparing text based on its underlying meaning rather than its surface form, which is useful in retrieving documents that are contextually related to a query. See
Chapter 11
for more information.
B. Allocating budget among different marketing channels is a scenario where the one-armed bandit problem can be used to model the decision-making process, as it involves choosing how to distribute resources among various options (marketing channels) with unknown outcomes. See
Chapter 12
for more information.
A. Object detection and recognition are fundamental in facial recognition applications, as they involve identifying and classifying faces into predefined categories. See
Chapter 12
for more information.
D. This scenario represents a resource allocation problem, where the objective is to distribute limited resources (water for irrigation) among competing activities or projects while adhering to constraints. See
Chapter 12
for more information.
Chapter 1What Is Data Science?
THE COMPTIA DATAX EXAM OBJECTIVES COVERED IN THIS CHAPTER INCLUDE:
Domain 4: Operations and Processes
4.1 Explain the role of data science in various business functions.
4.5 Given a scenario, implement best practices throughout the data science life cycle.
Domain 5: Specialized Applications of Data Science
5.4 Explain the purpose of other specialized applications in data science.
The rapid advances in data science have changed the way we work, live, and interact with the world around us. But what exactly is data science? Is it the same thing as machine learning? What about artificial intelligence? In this chapter, we define what data science is and how it differs from other closely related but distinct disciplines. We then explore some common applications of data science to a wide variety of problems in different domains. The chapter wraps up with a spotlight on data science best practices, which include the use of standardized workflow models and toolkits.
Data Science
Data science is an interdisciplinary field that has rapidly evolved to become a cornerstone of modern business, research, and technology. It encompasses a wide range of techniques and methodologies aimed at extracting meaningful information from both structured and unstructured data. The emergence of data science as a distinct discipline can be attributed to the digital revolution of the 21st century, which has led to an exponential growth in the volume, velocity, and variety of data. This deluge of data, often referred to as “big data,” presents both challenges and opportunities. The challenge lies in the ability to manage, process, and analyze vast amounts of data efficiently. The opportunity, on the other hand, is the potential to uncover hidden patterns, correlations, and insights that can inform strategic decisions, optimize processes, and create value.
At its core, data science integrates principles from statistics, mathematics, computer science, and domain-specific knowledge to unlock insights that can drive decision-making and innovation. Statistics and mathematics provide the foundational framework for data analysis, enabling data scientists to summarize data, test hypotheses, and draw inferences. Computer science, particularly in areas such as algorithms, data structures, database management, and programming, is essential for handling and processing data efficiently. Domain expertise, meanwhile, is crucial for understanding the context of the data and interpreting the results in a meaningful way.
One of the key strengths of data science is its applicability across a wide range of domains. In healthcare, data science is used to develop predictive models for disease outbreaks, personalize treatment plans, and improve patient outcomes. In finance, it is applied to detect fraudulent transactions, manage risk, and optimize investment strategies. Retailers use data science to understand customer behavior, forecast demand, and enhance the shopping experience. The applications are virtually limitless, spanning sectors such as manufacturing, education, transportation, and government.
As data continues to play an increasingly central role in society, the importance of data science cannot be overstated. It has the potential to drive innovation, improve efficiency, and solve complex problems in virtually every area of human endeavor. The field of data science is not only a fascinating area of study but also a critical driver of progress in the modern world.
Data Science, Machine Learning, and Artificial Intelligence
The term “data science” is frequently misunderstood and conflated with closely related but distinct fields such as machine learning and artificial intelligence. While these disciplines share some commonalities and often work in tandem, each has its own unique focus and scope. As shown in Figure 1.1, data science is an umbrella term that encompasses a broad range of techniques and methodologies for extracting knowledge and insights from data.
FIGURE 1.1 Data science, machine learning, and artificial intelligence
Data science encompasses the entire data processing lifecycle, including data collection, storage, cleaning, analysis, and visualization. It also involves using data analysis tools and techniques to inform business decision-making. Additionally, data science includes practices and policies to ensure ethical data use, regulatory compliance, and the protection of data privacy and security.
Artificial Intelligence
Artificial intelligence (AI) is a broad field that aims to create systems or machines that can perform tasks that typically require human intelligence. This includes reasoning, learning, problem-solving, perception, and language understanding. AI encompasses various techniques and approaches, including rule-based systems, expert systems, and machine learning.
Machine Learning
Machine learning is a subset of AI that focuses on developing algorithms that enable computers to learn from and make predictions or decisions based on data. It is one of the key approaches behind many AI applications, such as image recognition, natural language processing, and recommendation systems.
