Monte Carlo Tree Search for Real World Problems in Python

Name: Monte Carlo Tree Search for Real World Problems in Python
Rating: 4.6 (6 reviews)

Solve Business Problems with MCTS: Hands-on, and Spelled out Approach

Created byHadi Aghazadeh

Last updated 2/2025

English

What you'll learn

Fundamental Theory and Hands on Practice on Monte Carlo Simulation
Tree Data Structure
Tree Search Algorithms
Theory of Monte Carlo Tree Search
Hands-on Coding on Applying Monte Carlo Tree Search for Solving Job Shop Scheduling Problem
Learn How to Apply Monte Carlo Tree Search to Other Practical Real-world Problems

Course content

4 sections • 33 lectures • 7h 5m total length

Introduction8:10
Let's get started with the course, answering some of the most frequent and valid questions regarding the course and also let's talk about the course itself and why it is fundamental to learn MCTS.

Set up coding IDE4:19
We will set up coding environment for the whole course.
Monte Carlo Simulation: Theory17:30
We will go through the essential theories behind Monte Carlo Simulation.
Hands-on Practice with Monte Carlo Simulation in Python- Part 126:21
Let's create a fun experiment, to test the power of monte carlo simulation.
Hands-on Practice with Monte Carlo Simulation in Python - Part 218:36
Let's continue creating a fun experiment, to test the power of monte carlo simulation.
Results for Monte Carlo Simulation in Python4:19
This video is about obtaining results for monte carlo simulation and inferencing it.
You can download the notebook code in this section.

Introduction to Tree Data Structures11:00
Tree data structures are one of the most important data structures in computer science. We will learn its most important concepts.
Breadth First Search, and Depth First Search Algorithms: Theory14:07
Let's learn BFS and DFS algorithms which are the backbones for all other tree search algorithms.
Recursion Concept9:03
Understanding Recursion is very important for many sequential optimization problems. Let's learn it here.
Hands-on Code: Implement State Class23:22
We will implement the state class for solving a practical problem using BFS and DFS.
Hands-on Code: Implement State Transition15:14
Let's implement state transition part of the code.
Hands-on Code: Implement BFS Algorithm10:14
Finally we are ready to implement the first tree search algorithm: BFS.
Hands-on Code: Implement DFS Algorithm11:48
Let's implement BFS with recursive functionality as well.
Hands-on Code: Implement Main Loop20:11
Let's write the main code to obtain the results.
Hands-on Code: Obtain Results3:17
Results and the model comparisons will be presented here.
You can also download the code from this section.

MCTS Theory: Fundamentals14:58
Let's review fundamental theories behind MCTS and their origins.
MCTS Theory: Algorithm and Steps14:20
We will learn four steps of MCTS algorithm.
MCTS Theory: Numerical Example13:09
We will solve a small case of MCTS for few iterations by hand and with manual calculation.
Hands-on Code: Overall Schema for MCTS Coding8:25
Let's review all the parts will implement for MCTS and how they are related to each other.
Hands-on Code: Implement Rollout Module14:04
Let's start coding with rollout module which puts together all the steps in MCTS.
Hands-on Code: Implement Select and Expand Modules12:06
We will implement Select and Expand modules in this video.
Hands-on Code: Implement Simulation and Backpropagation Modules14:30
Let's finish Implementing required steps for MCTS.
Hands-on Code: Implement UCB and Greedy Action Selection14:40
Implementing of UCB and greedy algorithms here.
Hands-on Code: Implement Abstract Class for Node14:16
Let's build a template for how to define Node class for all the problems that can be solved with MCTS.
Hands-on Code: Define a Class for Job Shop Scheduling Problem11:53
Let's understand JSS data and define its class.
Hands-on Code: Implement Find Children Module13:14
Find children is crucial part of Node Class to find all possible children for current node. Let's implement it here.
Hands-on Code: Implement Find Random Child Module6:22
Let's Implement Find Random Child for the current node.
Hands-on Code: Implement Terminal and Reward Modules6:41
We are done with implementing all modules for JSS node class by finishing this video.
Hands-on Code: Obtain the First Results and Expriments20:58
We will obtain results and perform some experiments to obtain best possible parameters for MCTS.
Hands-on Code: Obtain Schedule11:28
We will obtain schedule in tabular mode here.
Hands-on Code: Implement Constraints Validation Function15:39
We need to check whether we meet constraints for JSS problem or not. Let's implement its function.
Hands-on Code: Implement Gantt Chart Function18:46
We will design a visualization function to see the obtained schedule in a better way.
Conclusion2:42
Let's wrap up the course and consider the next steps after this course.

You can download the code for MCTS and the data for JSS problem in this video as well.

Requirements

Basic understanding of Python programming language would be a great help.
No experience of Reinforcement Learning or any other optimization algorithms is needed. You will need all the required theory in this course.

Description

Unlock the power of Monte Carlo Tree Search (MCTS) and learn how to apply this cutting-edge algorithm to real-world business challenges! In this hands-on course, we’ll take you from the foundational theory of Monte Carlo simulations to advanced MCTS implementations, all in Python.

What makes MCTS truly practical is its versatility. Whether you're optimizing supply chain logistics, scheduling complex tasks, enhancing game AI, or making strategic business decisions under uncertainty, MCTS shines where traditional algorithms struggle. Its ability to balance exploration and exploitation makes it perfect for solving problems with large, dynamic, and unpredictable environments—just like in real-world business scenarios.

You’ll start with the basics—understanding Monte Carlo simulations and Python coding strategies. Then, we’ll dive deep into tree search algorithms like BFS and DFS, setting the stage for mastering MCTS. Through step-by-step coding sessions, you'll implement key MCTS components: rollout, selection, expansion, and backpropagation.

But we don’t stop at theory. You’ll solve practical business problems, including job shop scheduling, using MCTS with real-world data. We’ll guide you through designing code structures, optimizing performance, and analyzing results effectively.

By the end of this course, you'll not only understand how MCTS works but also how to apply it confidently to complex decision-making problems.

Who this course is for:

Applied Data scientists who are dealing with solving real-world problems.
Researchers who want to apply MCTS or combine their approach with MCTS.
Game developers who want to learn one of the most required algorithms for game development.
Operations research scientists who want to add new, yet powerful weapon to their optimization arsenal.
Planning and Scheduling Specialists who want to apply simple yet efficient algorithm to solve their daily complex planning tasks

Monte Carlo Tree Search for Real World Problems in Python

What you'll learn

Explore related topics

Course content

Introduction1 lecture • 8min

Monte Carlo Simulation5 lectures • 1hr 11min

Tree Search Algorithms9 lectures • 1hr 58min

Monte Carlo Tree Search Algorithm18 lectures • 3hr 48min

Requirements

Description

Who this course is for: