Prepare and test your skills

Prepare and test your skills

Free sample question with answers and detailed explanation below. The Cloud Digital Leader practice bank has an additional 3,070 practice questions.Try 10 free realistic questions: instant score, no credit card, and no account required.
A real estate organization has hundreds of thousands of property photos and wants to automatically categorize them by architectural style. The team plans to follow a foundational machine learning (ML) workflow on Google Cloud:
Which approach should the organization use for this workflow?
Supervised learning is a foundational machine learning approach in which a model learns to map inputs to known outputs by analyzing labeled training data. Each example in the training dataset consists of input features (such as the pixels of an image or columns in a table) paired with a corresponding ground-truth label or target value.
Unlike unsupervised learning (which ignores existing labels and only groups raw data by similarity) or reinforcement learning (which relies on sequential trial-and-error rewards in an interactive environment), supervised learning directly leverages the organization's pre-annotated examples to train a reliable classifier across the standard ML lifecycle.
Classical rule-based programming is the traditional software development paradigm in which a human programmer writes explicit, step-by-step instructions (conditional logic and deterministic rules) for a computer to follow. The system processes input data strictly according to these hard-coded rules to produce an output.
if-then rules to account for every pixel combination across hundreds of thousands of photos is practically impossible.This option would be appropriate if the organization wanted to categorize properties using simple, fixed business rules based on structured fields—such as labeling any building constructed before 1920 as "Historic" and any building with more than 10 floors as a "High-Rise."
A core reason organizations adopt machine learning over classical programming is to solve problems where rules cannot be easily codified by hand, allowing the system to learn patterns directly from example data.
Unsupervised learning is a machine learning paradigm where algorithms analyze unlabeled datasets to discover hidden structures, correlations, or natural groupings without human-provided answers or target labels. Instead of predicting a predefined category, the model explores the underlying distribution of the input data on its own.
This option would be the right choice if the organization had thousands of raw customer records or property records with no predefined labels and wanted to discover natural customer segments or unexpected groupings in the data without knowing the categories ahead of time.
Organizations sometimes confuse unsupervised learning with supervised classification because both organize items into groups. However, when predefined categories and labeled training examples exist, supervised learning is the appropriate approach.
Reinforcement learning is a machine learning approach in which an autonomous agent learns to make a sequence of decisions by interacting with a dynamic environment. Instead of being trained on a static dataset of labeled examples, the agent performs actions, observes the resulting state of the environment, and receives numerical rewards or penalties as feedback.
This approach would be appropriate if the organization were training an automated smart-building HVAC control system to dynamically adjust heating and cooling settings throughout the day based on sensor feedback to minimize energy costs while maintaining occupant comfort.
While supervised learning, unsupervised learning, and reinforcement learning are the three primary branches of machine learning, image categorization from labeled examples is a classic supervised learning task rather than an agent-based control problem.
Answer them yourself and get your score the moment you finish. No account needed.
Full access includes
Keep the momentum going with these hand-picked practice scenarios
Last updated