Table of Contents
Chapter 1 - Foundations of Artificial Intelligence1
| Introduction | 1 |
| Perception: Machines That See and Hear | 1 |
| Learning: The Engine of Improvement | 2 |
| Reasoning: Drawing Conclusions from Knowledge | 2 |
| Decision-Making: From Insight to Action | 3 |
| Machine Learning: The Core of Modern AI | 4 |
| Deep Learning | 5 |
| Key Model Architectures and Algorithms | 6 |
| The AI Model Development Lifecycle | 7 |
| Model Availability and Reliability | 9 |
| AI in Practice: Applications and Examples | 10 |
| Key Considerations in Building AI Systems | 12 |
| Summary | 13 |
| Review Questions | 14 |
Chapter 2 - Understanding Data: The Raw Material of AI15
| Introduction | 15 |
| The Importance of Data in AI | 15 |
| Types of Data in AI | 16 |
| Data Sources and Collection Techniques | 16 |
| The Dimensions of Data Quality | 18 |
| Data Labeling and Annotation | 19 |
| Data Augmentation and Synthetic Data | 21 |
| Data Pipelines and Infrastructure: From Lakes to Warehouses | 22 |
| Well-Known Datasets and Their Impact on AI | 24 |
| Legal and Ethical Considerations | 25 |
| Summary | 26 |
| Review Questions | 28 |
Chapter 3 - Machine Learning Fundamentals29
| Introduction | 29 |
| Key Machine Learning Model Types | 29 |
| Decision Trees | 29 |
| Ensemble Methods: Random Forests and Boosting | 30 |
| Bagging and Random Forests | 30 |
| Boosting and Gradient Boosting | 31 |
| Neural Networks and Deep Learning | 31 |
| Convolutional Neural Networks | 32 |
| Recurrent Neural Networks | 32 |
| Transformer Models | 33 |
| Fundamental Training Concepts | 34 |
| Overfitting and Underfitting | 34 |
| Bias-Variance Tradeoff | 35 |
| Setting the Learning Rate | 36 |
| Early Stopping | 36 |
| Model Evaluation Metrics | 37 |
| Summary | 37 |
| Review Questions | 39 |
Chapter 4 - Natural Language Processing40
| Introduction | 40 |
| Key Concepts in NLP Modeling | 40 |
| Tokenization and Text Preprocessing | 40 |
| Text Vectorization | 41 |
| Model Training and Fine-Tuning | 41 |
| Model Evaluation | 42 |
| Deployment and Machine Learning Operations | 42 |
| NLP Applications Across Industries | 44 |
| Finance | 44 |
| Healthcare | 45 |
| Customer Support | 46 |
| Legal | 48 |
| Retail | 50 |
| Challenges and Best Practices in Building NLP Systems | 51 |
| Summary | 52 |
| Review Questions | 53 |
Chapter 5 - Computer Vision54
| Introduction | 54 |
| Image Processing Fundamentals | 54 |
| Image Classification | 55 |
| Object Detection | 55 |
| Image Segmentation | 56 |
| Video Analysis and Temporal Models | 56 |
| Deployment of Vision Models | 57 |
| Common Challenges in Computer Vision | 58 |
| Summary | 59 |
| Review Questions | 61 |
Chapter 6 - The Responsible Use of AI62
| Introduction | 62 |
| The Role of Ethics in the AI Development Lifecycle | 62 |
| Algorithmic Bias: Sources and Risks | 62 |
| Principles of Fairness and Inclusion | 63 |
| Transparency and Explainability | 64 |
| Human Oversight and Accountability | 64 |
| Governance Mechanisms in AI Teams and Organizations | 65 |
| Long-term Responsibility and Public Trust | 65 |
| Summary | 66 |
| Review Questions | 67 |
Answers to Chapter Questions68
Glossary73
Index75
Course Details
Author: Steven M. Bragg, CPA
Steven Bragg, CPA, has been the chief financial officer or controller of four companies, as well as a consulting manager at Ernst & Young. He received a master’s degree in finance from Bentley College, an MBA from Babson College, and a Bachelor’s degree in Economics from the University of Maine. He has been a two-time president of the Colorado Mountain Club, and is an avid alpine skier, mountain biker, and certified master diver. Mr. Bragg resides in Centennial, Colorado. He has written more than 300 books and courses, including New Controller Guidebook, GAAP Guidebook, and Payroll Management.
Publication/Revision Date: 12/15/2025
Course Exam Questions (online): 25 (multiple-choice)
Program Delivery Method: NASBA QAS Self-Study
Available Formats of Course Text: PDF or PDF plus printed copy sent in the mail
Course Level, Prerequisites, and Advance Preparation Requirements
| License | Course Level | Prerequisites | Advance Preparation Requirements |
|---|
| CPA | Overview | None | None |
* This program is appropriate for professionals at all organizational levels.
Sponsor ID Numbers
National Registry of CPE Sponsors ID: 107615
State CPA Board Sponsor ID Numbers (where applicable)
Florida Division of Certified Public Accounting: 0004761
Hawaii Board of Public Accountancy: 14003
New York State Board for Public Accountancy: 002146
Ohio Accountancy Board: CPE .51 PSR
Pennsylvania State Board of Accountancy: PX178025
Texas State Board of Public Accountancy: 009349
Learning Objectives
As a result of studying the course material, you should be able to meet the objectives listed below:
- Identify the best uses for different types of neural networks.
- Specify the steps included in the machine learning project lifecycle.
- Describe the differences between supervised and unsupervised learning.
- Identify the drawbacks of deep learning models.
- Specify the different types of data used to train AI models.
- Identify the different types of data augmentation.
- Recall the differences between overfitting and underfitting.
- Specify the solutions to overfitting and underfitting.
- Describe the types of metrics that can be used to evaluate AI models.
- Specify why practitioners prefer ensemble methods over single decision trees.
- Specify how text vectorization works in natural language processing.
- Recall the different types of text preprocessing steps.
- Recall the advantages and disadvantages of word-level and character-level tokenization.
- Specify the best uses for a convolutional neural network (CNN).
- Identify why a CNN uses a convolutional filter to process an image.
- Recall how the optical flow concept applies to computer vision.
- Recall how the concept of data poisoning impacts machine learning security.
- Specify how the use of proxies can lead to indirect discrimination.
- Identify the reasons supporting the use of data anonymization.
- Specify the actions that can be taken to protect user data privacy in AI development.
