Intelligent Optimization

Optimization meets Machine Learning
Roberto Battiti, Kevin Tierney and Mauro Brunato

Cartoons are courtesy of Marco Dianti.

The strong coupling between Mathematical Optimization and Machine Learning pioneered by LION is rapidly gaining momentum in different research and business projects.

This book presents two topics which are in most cases separated: Machine Learning (the design of flexible models from data) and Intelligent Optimization (the automated creation and selection of optimal or improving solutions). More and more innovative people can now master the source of power arising from LION techniques to solve problems, improve businesses, create new applications.

Powerful tools are not only for cognoscenti and this book does a serious effort to distinguish the paradigm shift brought about by machine learning and intelligent optimization methods from the fine details, and it does not refrain from presenting concrete examples and vivid sticky images (made to stick to your mind).

Get the IObook

The book is now available in draft form (pdf) for personal and non-profit usage.
Write down your email address if you would like to get your personal copy. Please allow for some processing time by human volunteers (the delivery is not automated yet).

My name:
My email address:
My institution: (optional)
	I consent to the storage and treatment of my email address to receive information about Intelligent Optimization book and related activities. We will not share email addresses with third parties.

Please cite our book as:
Roberto Battiti, Kevin Tierney, Mauro Brunato
Intelligent Optimization — Optimization plus Machine Learning for Reliable Artificial Intelligence.
LION association, Italy, 2026.
[ BibTeX ]

Tutorials

The book will be accompanied by the periodic release of tutorials about topics related to intelligent optimization. The tutorials, written in Python, are presented as static web pages; however, the source Jupyter notebooks are provided to encourage readers to explore different scenarios and test their own ideas.

So far, the available topics are:

Combinatorial Optimization

You will be guided through a series of tentative solutions of the Vehicle Routing Problem, from simple greedy constructions to more and more effective ways to refine and improve them.

Greedy Algorithms, part 1: introduction to the Vehicle Routing Problem and to building simple solutions.
Greedy Algorithms, part 2: refining a constructive heuristic.
Local Search: improving existing solutions via small changes.
Coming next — repeated local search: continued solution improvement with restarts.

Intelligent Optimization

Problems at the crossroads of Machine Learning and Optimization Heuristics: a problem will be introduced and a solution based on Intelligent Optimization principles will be developed step by step.

Learning to park a car: a control task carried out by a neural network trained by a continuous oprimization algorithm.
Coming next — Parking a car behind another car: new constraints for our control task.