An agent can improve its performance by

Answer: An agent can improve its performance by learning.

Why: Learning means updating an agent's knowledge, parameters, or decision rules based on experience or feedback so that future actions lead to better task performance.

• Learning transforms observations and outcomes into improved behavior.

• Perceiving and observing supply the data an agent needs, and responding is how the agent acts, but none of these by themselves change the agent unless learning occurs.

• Example: In reinforcement learning, the agent learns from rewards to choose actions that increase future reward, improving performance over time.

Practical steps an agent uses to improve:

1. Collect data by perceiving or observing the environment.

2. Use the data and feedback to update internal models or policies (learning).

3. Apply the improved policy when responding so future performance is better.