An RL agent is trained to navigate a maze. The reward is +100 for reaching the exit and -1 for each step. The agent learns a policy that reaches the exit but takes a long, winding path. Which of the following is the most likely reason?

The reward for reaching the exit is too low compared to the step penalty.

The exploration rate is too low, preventing the agent from finding the shortest path.

The learning rate is too high, causing the agent to overreact to recent rewards.

The discount factor is too low, causing the agent to undervalue future rewards.

Which of the following is a common application of reinforcement learning?

Predicting house prices based on historical data.

Training a model to classify emails as spam or not spam.

Teaching an autonomous vehicle to navigate by rewarding safe driving behaviors.

Clustering customer data to identify market segments.

A recommendation system uses RL to suggest articles. Initially, it shows mostly random articles to learn user preferences, but over time it increasingly shows articles that have received high engagement. This approach is an example of:

Supervised learning, because the system learns from user engagement labels.

A trade-off between exploration and exploitation, balancing learning new preferences with using known preferences.

Exploitation, because the system is using known good articles.

Exploration, because the system is trying new articles to gather information.

In a reinforcement learning system for a robot navigating a warehouse, which of the following are components of the RL framework? (Select all that apply.)

The reward signal, such as +1 for reaching a destination.

The robot's current location and battery level as the state.

The robot's movement commands as actions.

A policy that maps states to actions.

A labeled dataset of optimal paths.

What is a 'policy' in reinforcement learning?

A policy is the set of all possible rewards in the environment.

A policy is the value of being in a particular state.

A policy is a function that maps states to actions, defining the agent's behavior.

A policy is the environment's response to the agent's actions.

Certified Tester AI Testing

Domain 3Objective 3

Reinforcement Learning CT-AI Practice Questions (Page 3)

Part of the Domain 3: Machine Learning (ML) - Overview domain, which makes up ~15% of our current practice bank. ISTQB does not publish an official question count, but from its 60-minute exam (~25–40 total, ~4–6 in this domain), expect 1–1 from this objective — we provide 24 practice questions to prepare you well beyond it. (estimate)

24questions here

5free pages

6concepts

Questions 11–15

11expert · hard
An RL agent is trained to navigate a maze. The reward is +100 for reaching the exit and -1 for each step. The agent learns a policy that reaches the exit but takes a long, winding path. Which of the following is the most likely reason?
Select an answer first
12foundation · easy
Which of the following is a common application of reinforcement learning?
Select an answer first
13application · medium
A recommendation system uses RL to suggest articles. Initially, it shows mostly random articles to learn user preferences, but over time it increasingly shows articles that have received high engagement. This approach is an example of:
Select an answer first
14application · medium · select all that apply
In a reinforcement learning system for a robot navigating a warehouse, which of the following are components of the RL framework? (Select all that apply.)
Select an answer first
15foundation · easy
What is a 'policy' in reinforcement learning?
Select an answer first

Finished these 5 questions?

Review the revealed explanations, or continue through the curriculum.

Next 5 questions →Back to CT-AI

Free Basic Practice is a study aid with revealable answers — not a scored exam. Examers.io is independent and not affiliated with or endorsed by ISTQB. “CT-AI” is a trademark of its owner, used for identification only.