MCQs

Disclaimer:

Educational Use Only:

The website links provided are to be used strictly for educational and knowledge-gaining purposes only.

No Misuse:

Do not misuse the links or the information contained therein for any commercial, illegal, or unethical activities.

Respect for Websites and Businesses:

Respect the intellectual property and business interests of the websites and their owners. Do not attempt to alter, copy, or redistribute content without proper authorization.

Acknowledge Value of Knowledge:

Recognize and value the knowledge and resources shared by these websites. Use the information responsibly and ethically.

No Endorsement:

The provision of these links does not imply endorsement of the linked sites, their content, or their products and services. Users are responsible for evaluating the suitability and accuracy of the information.

User Responsibility:

By accessing these links, you agree to use the information at your own risk and to comply with all applicable laws and website terms of use.

Overall MCQs for Reference

What does the discount factor (γ) in reinforcement learning primarily control?
a. Agent’s exploration rate
b. Importance of future rewards
c. Reward function shape
d. State transition model
Answer: b. Importance of future rewardseasyexamnotes+3
If the discount factor is set close to 0, how does the agent behave?
a. Ignores immediate rewards
b. Prioritizes immediate rewards
c. Focuses only on penalties
d. Seeks random actions
Answer: b. Prioritizes immediate rewardsmilvus+2
Which value of the discount factor makes the agent most farsighted?
a. 0
b. 0.1
c. 0.9
d. 1
Answer: d. 1geeksforgeeks+2
What happens when the discount factor is set to 1 in an infinite-horizon task?
a. The agent disregards future rewards
b. The value may diverge
c. Only immediate rewards influence action
d. Learning stops
Answer: b. The value may divergezilliz+2
In policy iteration for reinforcement learning, what role does discounting play?
a. Helps convergence
b. Slows down updates
c. Removes cumulative reward
d. Disables exploration
Answer: a. Helps convergencemilvus+2
Which mathematical formula describes discounted reward in RL?
a.
b.
c.
d.
Answer: a. simplilearn+2
A low discount factor is suitable for tasks where:
a. Only delayed rewards exist
b. Immediate rewards are critical
c. The agent never receives feedback
d. The reward is always negative
Answer: b. Immediate rewards are criticalgeeksforgeeks+2
What is the effect of increasing the discount factor in a sparse reward environment?
a. Faster learning
b. Propagates value of distant rewards
c. More randomness
d. Ignores policy gradients
Answer: b. Propagates value of distant rewardseitca+2
In which range does the reward discount factor ( ) typically fall?
a. -1 to 0
b. 0 to 1
c. 1 to 10
d. 10 to 100
Answer: b. 0 to 1zilliz+2
What may happen if the discount factor is too low in practice?
a. Agent becomes farsighted
b. Agent ignores immediate feedback
c. Learning becomes myopic
d. State value increases
Answer: c. Learning becomes myopicsimplilearn+2
The sum of discounted rewards forms the:
a. Q-value
b. Value function
c. Policy
d. Reward function
Answer: b. Value functionmilvus+2
Which notation is commonly used for the reward discount factor?
a. Alpha ( )
b. Beta ( )
c. Gamma ( )
d. Sigma ( )
Answer: c. Gamma ( )geeksforgeeks+2
Which factor influences how much future reward is considered in RL?
a. Exploration rate
b. Discount factor
c. Reward shaping
d. State transition
Answer: b. Discount factoreasyexamnotes+2
The discounted sum of future rewards is called:
a. Accumulated error
b. Expected return
c. Greedy action
d. Temporal difference
Answer: b. Expected returneasyexamnotes+2
Why would you select a discount factor close to 1?
a. For short-term objectives
b. When long-term rewards are crucial
c. To avoid overfitting
d. To reduce state space
Answer: b. When long-term rewards are crucialzilliz+2
If a reward at time is 10 and , what is its present value?
a. 9
b. 8.1
c. 10
d. 7.5
Answer: b. 8.1simplilearn+1
What does a discount factor of 0 imply?
a. Agent ignores future rewards
b. All rewards are negative
c. Learning is maximized
d. Rewards are unpredictable
Answer: a. Agent ignores future rewardsmilvus+2
Discounting helps prevent:
a. Overestimation of future rewards
b. Immediate reward maximization
c. Action selection
d. Policy loss
Answer: a. Overestimation of future rewardseitca+2
The discount factor appears in which RL algorithm?
a. Decision trees
b. Supervised learning
c. Q-learning
d. Clustering
Answer: c. Q-learninggeeksforgeeks+2
What role does discounting play in Markov Decision Processes (MDPs)?
a. Ignores uncertainty
b. Adjusts future reward values
c. Increases sample size
d. Disables learning
Answer: b. Adjusts future reward valueszilliz+2
If , how much is the immediate reward valued compared to next step?
a. Equally
b. 5% less
c. 95% less
d. 50% more
Answer: b. 5% lesssimplilearn+2
Discount factor selection is most important when:
a. The agent takes random actions
b. Balancing short- and long-term rewards
c. Maximizing the learning rate
d. Reducing the reward space
Answer: b. Balancing short- and long-term rewardseitca+2
In RL terminology, usually represents:
a. Immediate reward
b. Discounted sum of future rewards
c. State transition
d. Penalty signal
Answer: b. Discounted sum of future rewardsgeeksforgeeks+2
Delayed rewards are made less significant by:
a. Lowering the learning rate
b. Decreasing the discount factor
c. Increasing reward magnitude
d. Random state transitions
Answer: b. Decreasing the discount factoreitca+2
Why is it important not to set the discount factor exactly to 1 for infinite-horizon problems?
a. The agent ignores all rewards
b. Value functions may diverge
c. Learning rate collapses
d. Agent becomes random
Answer: b. Value functions may divergemilvus+2
Discounted rewards help in:
a. Speeding up convergence
b. Penalizing bad actions only
c. Eliminating exploration
d. Removing need for feedback
Answer: a. Speeding up convergencezilliz+2
In the Q-learning equation, where does the discount factor appear?
a. As a penalty
b. Multiplying the maximum Q-value of the next state
c. With the learning rate
d. As an offset
Answer: b. Multiplying the maximum Q-value of the next statesimplilearn+2
Which environment would benefit least from a high discount factor?
a. Sparse reward environments
b. Fully observable games
c. Volatile, highly uncertain environments
d. Long-horizon tasks
Answer: c. Volatile, highly uncertain environmentseitca+2
In Deep Q-Networks, the choice of discount factor:
a. Is irrelevant
b. Influences how future rewards are treated
c. Changes input layer size
d. Only affects final score
Answer: b. Influences how future rewards are treatedmilvus+2
Which real-world scenario is analogous to a low discount factor?
a. Planning for retirement
b. Consuming all food immediately
c. Investing in stocks
d. Delayed gratification
Answer: b. Consuming all food immediatelygeeksforgeeks+2
The choice of discount factor is an example of:
a. Hyperparameter tuning
b. Online learning
c. Reward shaping
d. Automatic differentiation
Answer: a. Hyperparameter tuninggeeksforgeeks+2
Discounting is not relevant in environments where:
a. Immediate rewards dominate performance
b. Rewards are only delivered later
c. Sequential decision-making is critical
d. Planning is long-term
Answer: a. Immediate rewards dominate performancezilliz+2
The following best explains the theoretical reason for discounting rewards:
a. Uncertainty about future rewards
b. Computational convenience
c. Encourages exploration
d. Reduces error in state values
Answer: a. Uncertainty about future rewardseitca+2
In practical RL, adjusting the discount factor helps to:
a. Balance computational expense
b. Tune the agent’s prediction horizon
c. Set learning rate
d. Increase training data
Answer: b. Tune the agent’s prediction horizonmilvus+2
Without discounting, the expected return in infinite-horizon MDPs would:
a. Be unaffected
b. Be undefined or infinite
c. Simplify computation
d. Accelerate learning
Answer: b. Be undefined or infinitesimplilearn+2
Which equation describes the discounted value in RL?
a.
b.
c.
d.
Answer: a. geeksforgeeks+2
The use of discount factor in reward calculation can be attributed to:
a. Markov property
b. Bellman equation
c. Linear regression
d. Clustering property
Answer: b. Bellman equationsimplilearn+2
The main tradeoff in selecting the discount factor is between:
a. Risk and reward
b. Immediate and future rewards
c. Exploration and exploitation
d. Learning and planning
Answer: b. Immediate and future rewardseitca+2
What happens to agent’s policy if only future rewards are considered ( close to 1, rewards delayed)?
a. The agent learns quickly
b. The agent can hesitate long for larger rewards
c. The agent takes random actions
d. The value function decreases
Answer: b. The agent can hesitate long for larger rewardszilliz+2
The RL value function with discounting is used to:
a. Forecast state transitions
b. Estimate expected cumulative rewards
c. Tune action spaces
d. Maximize immediate penalty
Answer: b. Estimate expected cumulative rewardsmilvus+2
A zero discount factor simplifies RL to:
a. A supervised learning task
b. A greedy reward maximization
c. Policy iteration
d. Temporal difference learning
Answer: b. A greedy reward maximizationgeeksforgeeks+2
Why might discounting be omitted in episodic tasks with finite time horizons?
a. Rewards are negative
b. All episodes end quickly
c. Sum remains finite
d. State space is small
Answer: c. Sum remains finitezilliz+2
For a discount factor , what is always true?
a. Rewards decay exponentially over time
b. Rewards grow quadratically
c. No future rewards count
d. Rewards remain unchanged
Answer: a. Rewards decay exponentially over timesimplilearn+2
In the Bellman optimality equation, the discount factor:
a. Determines the influence of next state values
b. Adds a penalty
c. Adjusts learning rate
d. Removes exploration
Answer: a. Determines the influence of next state valuesmilvus+2
Discounting can be visualized as:
a. Stretching time horizon
b. Diminishing rewards into the future
c. Penalizing state transitions
d. Exploding values
Answer: b. Diminishing rewards into the futuregeeksforgeeks+2
The practical effect of discount factor in learning curves is to:
a. Control the speed of policy updates
b. Smooth the convergence path
c. Flatten the model
d. Only affect exploration
Answer: b. Smooth the convergence pathzilliz+2
In Monte Carlo methods, discounting:
a. Is ignored
b. Is essential for infinite episodes
c. Prevents early termination
d. Increases randomness
Answer: b. Is essential for infinite episodesmilvus+2
Discounted reward is a form of:
a. Present value computation
b. Random reward shaping
c. Supervised loss
d. Non-parametric update
Answer: a. Present value computationsimplilearn+2
What is one limitation of using a high discount factor?
a. Ignores short-term gains
b. Slows adaptation to immediate changes
c. Requires less memory
d. Reduces sample complexity
Answer: b. Slows adaptation to immediate changeseitca+2
The discount factor is most often chosen through:
a. Analytical derivation
b. Trial and error or experimentation
c. Default software settings
d. Genetic algorithms
Answer: b. Trial and error or experimentationgeeksforgeeks+2

Blog

Systemethic Technologies

Leave a Reply Cancel reply

Useful links

Other Terms & Policies

Contact Info