Microsoft Azure AI Fundamentals (AI-900)Describe fundamental principles of machine learning on AzureEasy
A financial institution is developing a machine learning model to detect potential money laundering activities. The model needs to analyze large volumes of transaction data to identify patterns that deviate significantly from normal behavior without prior labeled examples of money laundering. Which type of machine learning approach is most suitable for this task?
- AReinforcement Learning
- BSupervised Learning
- CSemi-supervised Learning
- DUnsupervised Learning
Show answer & explanationAnswer & explanation
Correct answer: D. Unsupervised Learning
The key phrase 'without prior labeled examples of money laundering' indicates that the model must learn from unlabeled data. Unsupervised learning is designed for this, identifying patterns, structures, or anomalies in data without human-provided labels.
Why the other options are wrong
- A. Reinforcement learning involves an agent learning through trial and error with rewards, not suitable for pattern detection in static transaction data.
- B. Supervised learning requires labeled data (i.e., examples of money laundering and legitimate transactions).
- C. Semi-supervised learning uses a small amount of labeled data with a large amount of unlabeled data, but the scenario specifies 'without prior labeled examples'.
Unsupervised Learning
Unsupervised learning is a type of machine learning where the algorithm learns from unlabeled data, identifying patterns, structures, or relationships within the data without explicit guidance.
- Deals with unlabeled data.
- Common tasks include clustering, dimensionality reduction, and anomaly detection.
- Aims to discover hidden structures in data.
Memory trick: Unsupervised: The machine learns 'unsupervised', finding its own patterns in the data.