In the past few years, the area of Machine Learning has witnessed tremendous advancements and achievements, becoming a pervasive technology in a wide range of applications, from industrial domains to everyday-used apps. One area that can significantly benefit from the use of machine learning is combinatorial optimization. Modeling and solving, the two pillar activities for dealing with Constraint Satisfaction and Optimization Problems, can both Machine Learning techniques for boosting their accuracy, efficiency and effectiveness.
In this tutorial we will show how Machine Learning techniques can be used to support the modeling activity by providing model components through machine learning; and to boost the search effectiveness, by understanding which parts of the search space are more promising, or by selecting the most suitable algorithm for a given problem from a portfolio of techniques. Connections with model predictive control and black box optimization will also be covered.
The tutorial will be composed of a talk, covering related recent literature and a hands-on session where some integration techniques will be explained and used.
We have now setup a web site to collect material related to Empirical Model Learning, a technique that allows embedding an ML model within an optimization model.
The web site is available at:
It contains:
You are invited to head there to get more information!
Talk:
Hands-on session (about Empirical Model Learning)
The tutorial is mainly addressed to: (1) researchers in decision support systems, constraint reasoning and optimization, (2) researchers in Machine Learning, (3) practitioners interested in applications where data are available and decision support systems are needed (examples, industrial applications, smart cities, mobility, energy, etc.).
The needed background knowledge concerns decision support systems, the main machine learning techniques, but the tutorial will provide also an introduction on the main techniques used and the concepts underlying them. Some knowledge of the Python programming language is required for the hands-on session.
Michele Lombardi is a fixed-term Assistant Professor at the DISI department of the University of Bologna; his PhD thesis on hybrid methods for resource allocation and scheduling won the AIxIA Marco Cadoli award in 2010, and honorable mentions ad CP2011 and ICAPS 2012. He works on Combinatorial Optimization and Decision Support Systems. In particular, his research activity is focused on hybrid optimization methods, based on heterogeneous techniques such as Constraint Programming, (Mixed) Integer Linear (and Non-Linear) Programming, and Machine Learning. His main application fields are Resource allocation and Scheduling problems, Off-line/on-line optimization and optimization under uncertainty, and optimization on complex systems. He is member of the editorial board of the Constraints Journal, and a co-founder of the MindIT AI start-up in Italy.
Michela Milano is full professor at the Department Computer Science and Engineering of the University of Bologna. She received her Ph.D. in Computer Science in 1998. Her research interests cover the area of hybrid optimization, a multi-disciplinary field at the cross-road of computer science and applied mathematics, the integration of machine learning with optimization and decision support techniques, optimization for embedded system design, computational sustainability. She is Board member of the European Association of Artificial Intelligence (EurAI) and executive councillor of AAAI. She is Editor in Chief of the Constraint Journal, Area Editor of Constraint Programming Letters and Area Editor of INFORMS Journal on Computing. She is the recipient of the Google Faculty Research Award on DeepOpt: Embedding deep networks in Combinatorial Optimization. She has been the coordinator of the EU FP7 project e-POLICY - Engineering the POlicy making LIfe CYcle (2011-2014), aimed at integrating optimization and decision support techniques with social simulation and machine learning to help policy makers in their decision process, and participant of a number of FP7 and H2020 EU projects on mobility, energy and computing.
Michela Milano and Michele Lombardi have been working in the area of Machine Learning for optimization since 2010. They are the main proponent of the Empirical Model Learning Learning framework, one of the methods to integrate ML and optimization for problem modeling.