OPERNEX LEARNING

Learn Agentic AI

A structured, practical learning path from language model foundations to intelligent agentic systems.

6 Learning parts
453 Subtopics
415 Published lessons
Foundations / Introduction to Reinforcement Learning / The Markov Decision Process (MDP)
Lesson 102 of 415 6 min read
3.1.4

Transition Function and Reward Function

Learn how actions change the environment and how rewards provide feedback.

What is it?

The transition function describes how the environment moves from one state to another after an action. It is represented by P(s′|s,a). The reward function describes the immediate reward and is represented by R(s,a,s′).

Key points

1. P(s′|s,a) represents the probability of reaching s′ from s after action a. 2. Transitions can be deterministic or stochastic. 3. The reward function evaluates the outcome. 4. R(s,a,s′) represents the transition reward. 5. Rewards guide the agent toward desirable outcomes.

Example: A robot moving through a warehouse

Current state: Robot at position A. Action: Move right. Transition: Environment determines the next position. Reward: The robot receives a reward or penalty depending on the outcome.

Real-world application

Transition and reward functions model uncertainty and learning objectives.
Key takeaway

The transition function describes what happens next; the reward function evaluates the outcome.