Corporate Business Alliance

CBA-AIP · Study material

Contents

The complete contents of the CBA Certified Artificial Intelligence Professional study material: every chapter and lesson, the reading time of each, and where the assessed workbooks fall. One lesson is open to read in full.

Lessons

97

Reading time

5 hours

Workbooks

19

Words

63,414

These figures are counted from the material itself. The reading time assumes a steady pace and no re-reading, and does not include the time spent building the workbooks.

Free, no account

Read a complete lesson

What machine learning is, and what it is not, from AI and Machine Learning Foundations, as a candidate reads it, slides included.

Read the lesson

Contents

7 chapters, 97 lessons. Lessons marked workbook carry a spreadsheet you build and submit, marked against the published tolerance. The award requires 4 of the 19 completed to that tolerance.

Introduction

1 lesson · 2 min

  1. CBA-AIP Study Guide2 min

01AI and Machine Learning Foundations

15 lessons · 53 min

  1. What this chapter covers2 min
  2. What machine learning is, and what it is notFree3 min
  3. The three ways a system learns from data5 min
  4. Classification, regression, and the difference between a score and a decision2 min
  5. Features, labels, training and inference5 min
  6. Model families and what each is for2 min
  7. Judging a classifier: reading the matrix and then reading the moneyworkbook7 min
  8. Overfitting, underfitting and generalisationworkbook2 min
  9. Judging a forecast: the baseline is the whole argument5 min
  10. What a model can and cannot learn from data6 min
  11. Neural networks and deep learning, and when the complexity is not worth it4 min
  12. Generative models: fluent by construction, accurate only by design3 min
  13. Where a rule, or classical statistics, beats a modelworkbook3 min
  14. Common pitfalls2 min
  15. Chapter summary2 min

02Data Foundations for AI

18 lessons · 44 min

  1. What this chapter covers2 min
  2. Why data work dominates real projects2 min
  3. What kind of data you are dealing with2 min
  4. The quality dimensions and how each breaks a model3 min
  5. Missing values, outliers and duplicatesworkbook2 min
  6. Where labels come from and what they cost3 min
  7. Label noise and the ceiling it puts on performanceworkbook3 min
  8. Splitting data: training, validation and test2 min
  9. Leakage: the failure that destroys a project quietly4 min
  10. Representativeness and sampling biasworkbook2 min
  11. Class imbalance and what it does to what a model learns3 min
  12. Feature engineering at a conceptual level2 min
  13. The data behind a generative system3 min
  14. Data drift against concept drift2 min
  15. Personal data, minimisation and retention3 min
  16. Documenting a dataset so somebody else can trust it2 min
  17. Common pitfalls2 min
  18. Chapter summary2 min

03Model Development and Evaluation

17 lessons · 57 min

  1. What this chapter covers2 min
  2. From business decision to modelling problem3 min
  3. Choosing what to predict3 min
  4. Baselines: the number that makes a result mean somethingworkbook5 min
  5. The confusion matrix, worked properlyworkbook4 min
  6. Choosing a metric that matches the cost of each error3 min
  7. Thresholds and operating pointsworkbook5 min
  8. ROC, AUC and precision-recall curves2 min
  9. Measuring error in regression and forecasting4 min
  10. Validation design: splits, cross-validation and the test set5 min
  11. Labels, leakage and the data behind the metric4 min
  12. Reading performance honestly: imbalance, subgroups and uncertaintyworkbook2 min
  13. Evaluating systems that have no single right answer4 min
  14. Explaining a model's behaviour, and the limits of that explanation2 min
  15. When is a model good enough to deploy, and who decides4 min
  16. Common pitfalls3 min
  17. Chapter summary2 min

04Deployment, Operations and the AI Lifecycle

17 lessons · 49 min

  1. What this chapter covers2 min
  2. The deployment gap: why a model that passed testing still fails3 min
  3. Deployment patterns: where the answer has to be, and when3 min
  4. Degraded modes: what the system does with no model2 min
  5. Integration: the model is a component, not a product2 min
  6. The human in the loop, and the capacity that decides everything5 min
  7. Monitoring: system health is not model health6 min
  8. Drift: what changes, and how you find outworkbook5 min
  9. Deciding whether to retrainworkbook3 min
  10. Release mechanics: shadow, canary, champion-challenger, rollback2 min
  11. Versioning and the audit trail2 min
  12. Incident response when a model misbehaves2 min
  13. Operating generative systems: evaluation that keeps working3 min
  14. The total cost of an AI system, and why the model is the cheap partworkbook3 min
  15. Roles, ownership and the end of the lifecycle2 min
  16. Common pitfalls2 min
  17. Chapter summary2 min

05AI Application Patterns and Generative AI

14 lessons · 46 min

  1. What this chapter covers2 min
  2. Choosing the pattern before choosing the technology5 min
  3. What a language model actually does, and what it cannot do2 min
  4. Why it states wrong things confidently: four causes, four fixes4 min
  5. Prompting as a specification2 min
  6. Retrieval augmented generation: what it fixes and what it does notworkbook5 min
  7. Diagnosing the returns-window fault end to end3 min
  8. Fine-tuning, retrieval or prompting: choosing between them3 min
  9. Evaluating generative outputworkbook6 min
  10. Designing human review4 min
  11. Cost and latency as design constraintsworkbook3 min
  12. Untrusted content, tool use and taking actions2 min
  13. Common pitfalls2 min
  14. Chapter summary3 min

06Responsible AI, Governance and Risk

15 lessons · 49 min

  1. What this chapter covers2 min
  2. Responsibility is a property of the system, not of the model2 min
  3. Fairness: what it means concretely, and why the measures conflictworkbook6 min
  4. Where bias enters: data, labels, features and deployment6 min
  5. Choosing a threshold is a policy decision, not a technical one3 min
  6. Explaining a model and explaining a decision3 min
  7. Human oversight that is real rather than nominal6 min
  8. Privacy and personal data in AI systems3 min
  9. Risk classification and proportionate controlworkbook4 min
  10. Documentation an auditor or regulator would acceptworkbook4 min
  11. Accountability: who owns a model's decisions2 min
  12. Procurement of third-party AI2 min
  13. Security and misuse as governance problems2 min
  14. Common pitfalls2 min
  15. Chapter summary2 min

What the material is

The study material is written to the discipline’s domains and learning objectives, chapter by chapter, with worked examples and unscored checks in the lessons and assessed workbooks where the level requires applied work. It does not contain a bank of practice questions matching the examination; the specimen paper shows the style and standard of the examination, with a written rationale for every option.

Back to CBA-AIPCertification scheme