AML Cheat Sheet 2026

The 30 highest-yield AML facts, distilled from real exam questions. Print it, save it as a PDF, or study it here — free, no sign-up.

1. What effect does setting the discount factor (γ) close to 0 have on a reinforcement learning agent? → The agent becomes myopic, prioritizing only immediate rewards
2. What does the Bellman equation fundamentally express in reinforcement learning? → The recursive relationship between a state's value and the values of its successor states
3. What does a log transformation primarily help with during feature engineering? → Reducing the skewness of right-skewed distributions
4. Which neural network is best suited for sequential data? → Recurrent Neural Network
5. Which method identifies predictive features by measuring how much each reduces impurity in a trained Random Forest? → Feature importance from Random Forest
6. What does accountability in AI mean? → Ensuring responsibility for AI impacts
7. In AML practice, what is the purpose of a standard operating procedure (SOP)? → To document step-by-step instructions for routine tasks to ensure consistency and quality
8. Which NLP technique converts words into dense vector representations that capture semantic relationships? → Word2Vec / Word Embeddings
9. Which method involves searching through a range of hyperparameters? → Grid search
10. What does 'algorithmic bias' refer to? → Unfair outcomes produced by models
11. What is the purpose of max pooling in a convolutional neural network? → Reduce spatial dimensions while retaining dominant features
12. What is the BLEU score used to measure in NLP? → Quality of machine-generated text by comparing n-gram overlap with reference translations
13. Which type of learning does deep learning primarily fall under? → Supervised learning
14. What is the goal of data governance? → To manage data integrity and compliance
15. Which outlier detection method uses the interquartile range (IQR) to flag extreme values? → Tukey's fence method
16. What does professional liability insurance protect a AML practitioner against? → Financial loss from claims of negligence, errors, or omissions in professional services
17. What does professional liability insurance protect a AML practitioner against? → Financial loss from claims of negligence, errors, or omissions in professional services
18. What is the risk of using poor quality data in AI? → Incorrect and potentially harmful outcomes
19. What does the term 'overfitting' refer to in deep learning? → The model performs well on training data but poorly on unseen data
20. Which principle is essential for ethical AI? → Fairness
21. Which encoding technique assigns integer values to categories that have a natural order? → Ordinal encoding
22. What does a stochastic policy output for a given state in reinforcement learning? → A probability distribution over all possible actions
23. What does professional liability insurance protect a AML practitioner against? → Financial loss from claims of negligence, errors, or omissions in professional services
24. What is canary deployment in ML production systems? → Gradually rolling out a new model to a small subset of users before a full release
25. Which of the following is NOT a standard component of a Markov Decision Process (MDP)? → Neural network architecture
26. Which dimensionality reduction technique projects data onto directions of maximum variance? → PCA
27. Which technique prevents overfitting in optimization? → Regularization
28. In AML practice, what is the purpose of a standard operating procedure (SOP)? → To document step-by-step instructions for routine tasks to ensure consistency and quality
29. What is the purpose of model versioning in MLOps? → To track distinct model iterations with metadata to enable rollback and comparison
30. Which architecture is most commonly used for image-to-image tasks like semantic segmentation and medical image analysis? → U-Net (encoder-decoder with skip connections)