AIF-C01 Domain 1: AI and ML fundamentals

What to understand

Start with the difference between artificial intelligence, machine learning and deep learning. Then distinguish supervised, unsupervised and reinforcement learning by the information available during training—not by memorising product names.

• Classification predicts a category; regression predicts a number.
• Training learns parameters from data; inference applies the trained model to new input.
• Overfitting means strong training performance but weak generalisation.
• Accuracy, precision, recall and F1 answer different evaluation questions.

Common AIF-C01 decisions

Look for the business objective first. Fraud detection may value recall when missing fraud is costly; a workflow that sends cases to expensive human review may value precision. Data quality, representative samples and suitable evaluation metrics matter more than choosing the most complex model.