Databricks Certified Machine Learning Professional — Question 76

A machine learning engineer is working on a fraud detection machine learning application. When a transaction is made with a credit card, the machine learning application will immediately process the data and make a prediction to determine whether or not to approve the transaction based on the probability that the transaction is fraudulent.

Which deployment strategy can be used to meet these requirements?

Answer options

• A. Edge/on-device
• B. Real-time
• C. Batch
• D. Streaming

Correct answer: B

Explanation

The correct answer is B, Real-time, as it enables immediate processing and prediction of transactions, which is crucial for fraud detection. Options A (Edge/on-device) and C (Batch) do not provide the speed required for immediate decision-making, while D (Streaming) focuses on continuous data flow rather than instant predictions.