Lars Kotthoff

University of Wyoming

• 11/07/2023
• VHI 00.10 Aula Gaston Eyskens

Getting the Best out of your Constraint Solver: Portfolios and Automated Tuning

You've modeled your problem, now all you need to do is publish the results. But wait, the solver can't actually solve it! You're back to square one -- what to do? In this tutorial, we'll have a look at automated methods for getting better solving performance, without having to implement your own solver or fancy constraints. Using portfolios of solvers and automatically tuning their parameters, you can often get substantial performance improvements. We'll cover some of the basic methods (which are useful not only in Constraint Programming) and how to implement them in practice, with exercises that you can re-use for the constraints application that you're interested in. The methods are mostly agnostic to the particular constraint solvers and problems they are applied to and can be adapted quickly to other applications. We'll have a look at parallels to automated machine learning, current advances in this area, and open questions.

• Slides of the talk

Lars Kotthoff

University of Wyoming

• 11/07/2023
• VHI 00.10 Aula Gaston Eyskens

Lab session on Algorithm Portfolios and Automated Tuning

You will get hands-on experience with automated tuning tools that can help you to get the best performance out of CPMpy and other constraint programming solvers. We will have a look at the fundamental building blocks of state-of-the-art automated machine learning and how to use them in the context of CP.

• Google Colab that will be used

Elias Khalil

University of Toronto

• 11/07/2023
• VHI 00.10 Aula Gaston Eyskens

Predict-then-optimize: A tour of the state-of-the-art using PyEPO

In deterministic optimization, it is typically assumed that all parameters of the problem are fixed and known. In practice, however, some parameters may be a priori unknown but can be estimated from historical data. A typical predict-then-optimize approach separates predictions and optimization into two stages. Recently, end-to-end predict-then-optimize has become an attractive alternative. In this talk, I will provide an overview of the predict-then-optimize problem as well as a suite of end-to-end learning methods that have been developed independently in the operations research and machine learning communities. To ground the discussion, I will use the PyEPO package (https://github.com/khalil-research/PyEPO/), a PyTorch-based end-to-end predict-then-optimize library in Python. PyEPO (pronounced like 'pineapple' with a silent 'n') is the first generic tool for linear and integer programming with predicted objective function coefficients. It provides four base algorithms: a convex surrogate loss function from the seminal work of Elmachtoub & Grigas (2021), the differentiable black-box solver approach of Vlastelica et al. (2019), and two differentiable perturbation-based methods from Berthet et al. (2020). PyEPO provides a simple interface for the definition of new optimization problems, the implementation of state-of-the-art predict-then-optimize training algorithms, the use of custom neural network architectures, and the comparison of end-to-end approaches with the two-stage approach. PyEPO enables the first comprehensive comparison of these methods on problems such as Shortest Path, Multiple Knapsack, and the Traveling Salesperson Problem, including on an image dataset. The empirical insights therein could guide future research. Joint work with Ph.D. student Bo Tang at the University of Toronto.

• Slides of the talk

Elias Khalil

University of Toronto

• 11/07/2023
• VHI 00.10 Aula Gaston Eyskens

Lab session on Predict-then-optimizize: a tour using PyEPO

In this lab session, we will use the PyEPO library to get hands-on experience with predict-then-optimize problems.

• Google Colab that will be used

Everyone (optional)

• 11/07/2023
• Sports Centre KU Leuven (builde De Nayer)

Sport at the sports center

Let's get moving! Those that want can join us to stretch our legs. We will host beach volleyball and spikeball matches. Given enough interest, we can also host a 5km trail run in the nearby forests. There are other facilities as well, such as an athletic field, a running track (Finse piste), a basketbal field and more.

To get there, you can either walk or from the station take the 'Buitenringbus 600' to 'Leuven Sportcentrum' then walk to 'Gebouw De Nayer' (map).

Wouter Kool

ORTEC

• 12/07/2023
• VHI 00.10 Aula Gaston Eyskens

Deep Learning & Combinatorial Optimization

Combinatorial optimization is different from recognizing cats and dogs: in the limit of infinite computation, the task is trivial. Therefore, machine learning based approaches aim to reduce total computation, possibly at the expense of solution quality, by learning from past experience. A well-trained model can reduce the size of the relevant search space, but computation used for evaluating a model cannot be used for search, so this is a trade-off that should be carefully made: count your flops and make your flops count! In this talk I will go through practical examples in the context of this trade-off. I will present a number of challenges and provide some guidelines that may help define future research directions to apply deep learning for combinatorial optimization.

• Slides of the talk

Wouter Kool

ORTEC

• 12/07/2023
• VHI 00.10 Aula Gaston Eyskens

Learning to Improve a Hybrid Genetic Search Algorithm

In this lab, we will get our hands dirty in solving some actual optimization problems! We will dive into the hybrid genetic search (HGS) algorithm, a state-of-the-art algorithm for the vehicle routing problem with time windows (VRPTW). We will explore how we can parameterize an important part of the method using a deep neural network, to improve its efficiency. It is up to you to make it work!

• Google Colab that will be used

Michele Lombardi

Università di Bologna

• 12/07/2023
• VHI 00.10 Aula Gaston Eyskens

Rambling Away from Decision-Focused Learning

Combining of prediction and optimization has been an established practice in industrial decision support for a long time. In recent years, however, the topic has been the center of renewed research interest since the introduction of Decision-+Focused Learning (DFL). From a semantic point of view, DFL seeks to incorporate the decision-making phase at training time, so that predictions are optimized for minimum decision cost rather than for maximal accuracy. From a technical point of view, the key issue in DFL is finding ways to differentiate an argmin operator, so that optimization can be integrated with (stochastic) gradient descent. In this lecture, we will argue that the basic ingredients from DFL can be used for much more than its intended purpose. Starting from a technical review of DFL, we will start a 'ramble' that will lead to possibly unexpected places, including stochastic optimization, algorithm configuration, and reinforcement learning.

• Slides of the talk (PDF)
• Repository of the talk, with runnable notebooks

Everyone

• 12/07/2023
• Atrium, Campus Groep T

Poster session & Walking Dinner

Size Requirements: A1
Venue: Atrium, Campus Groep T. A beautiful location in the city center.
Event Setup: Poster session with walking dinner and drinks

Everyone (optional)

• 12/07/2023
• Oude Markt

Going for a drink on Oude Markt

Visiting the Oude Markt in Leuven means joining 'The longest bar in Europe'. Nowhere will you find more catering establishments together than on this unique square. No wonder that the Oude Markt is the meeting place of Leuven. Especially when the weather is nice, when the numerous terraces are filling up.
More info on the restaurants and bars in the center of Leuven can be found here.

Quentin Cappart

Polytechnique Montreal

• 13/07/2023
• VHI 00.10 Aula Gaston Eyskens

Learning a Value-Selection Heuristic Inside a Constraint Programming Solver

Constraint programming is known for being an efficient approach for solving combinatorial problems. Important design choices in a solver are the branching heuristics, which are designed to lead the search to the best solutions in a minimum amount of time.However, developing these heuristics is a time-consuming process that requires problem-specific expertise. This observation has motivated many efforts to use machine learning to automatically learn efficient heuristics without expert intervention. Although several generic variable-selection heuristics are available in the literature, the options for a generic value-selection heuristic are more scarce. In this talk, I will present a generic learning procedure that can be used to obtain a value-selection heuristic inside a constraint programming solver. This has been achieved thanks to a reinforcement learning algorithm combined with an architecture based on graph neural networks.

• Slides of the lecture

Quentin Cappart

Polytechnique Montreal

• 13/07/2023
• VHI 00.10 Aula Gaston Eyskens

Solving combinatorial problems with SeaPearl

The related talk presented a generic learning procedure for getting a value-selection heuristic inside a constraint programming solver. In this practical session, you will have the opportunity to test this idea with Seapearl, a CP solver based on reinforcement learning for selecting the next value to branch on. Seapearl is an open-source project, available here: https://github.com/corail-research/SeaPearl.jl

• Installation instructions for the lab session

Yingqian Zhang

TU Eindhoven

• 13/07/2023
• VHI 00.10 Aula Gaston Eyskens

Deep Reinforcement Learning in Solving Industrial Optimization Problem

Recently, many Deep Reinforcement Learning (DRL) approaches have been proposed to solve combinatorial optimization problems (COPs). Some approaches are end-to-end, which aim to learn the best decisions directly. Some are hybrid, i.e., learning is used to assist metaheuristics (search algorithms and evolutionary algorithms) in finding better decisions faster. I will discuss how to develop such learning based approaches to solve real-world industrial optimization problems. In addition, I will introduce the benchmark problem instances of machine scheduling that we have been developing for comparing different learning-based approaches.

• Slides of the talk

Tias Guns

KU Leuven

• 13/07/2023
• VHI 00.10 Aula Gaston Eyskens

Intro and start of Hackathon

During this hackathon we will tackle one of the open questions of any of the topics presented during the summer school. Every speaker at the summer school will propose an open topic and provide some guidance in how to get started. You will collaborate in groups of 3-5 people to come up with a clever implementation that relates to the problem. Each group will present their progress and potentially some results on Friday at noon.

• Doc with ideas and group formation

Everyone (optional)

• 13/07/2023
• Center of Leuven

Culture night: Museum M or brewery visit

On thursday, we propose you visit some of the musea Leuven has to offer. In particular, the Museum M or a brewery/local pubs (tickets not included). The museum M displays old and new art inspired by Leuven’s versatility. The collection is mainly focused on the art production in Leuven and Brabant from the Middle Ages to the 19th century.

Dimos Tsouros

KU Leuven

• 14/07/2023
• VHI 00.10 Aula Gaston Eyskens

Constraint Acquisition: Learning Constraint Models from Data

The basic assumption in CP is that the user models the problem, and a solver is then used to solve it. However, expressing a combinatorial problem as a constraint model is not always straightforward. As a result, modelling is considered to be a bottleneck for the wider use of CP. To overcome this obstacle, several techniques have been proposed for modeling a constraint problem (semi-)automatically. In constraint acquisition, the model of a constraint problem is acquired (i.e., learned) using a set of examples of solutions, and possibly non-solutions. The talk will be an overview of constraint acquisition research, in which learning techniques are used to learn constraint models from data. This is done either in a passive setting, using an existing set of solutions and/or non-solutions of the problem, or in an active setting where the system interacts with the user to model the problem. I will discuss both passive and (inter)active learning, the current state-of-the-art and the connections to the machine learning field. Finally, I will focus on the current challenges in Constraint Acquisition.

Everyone

• 14/07/2023
• VHI 01.40 Collaborative Room

Hackathon continuation

During this hackathon we will tackle one of the open questions of any of the topics presented during the summer school. Every speaker at the summer school will propose an open topic and provide some guidance in how to get started. You will collaborate in groups of 3-5 people to come up with a clever implementation that relates to the problem. Each group will present their progress and potentially some results on Friday afternoon.

Everyone

• 14/07/2023
• VHI 01.40 Collaborative Room

Hackathon student presentations

During this hackathon we will tackle one of the open questions of any of the topics presented during the summer school. Every speaker at the summer school will propose an open topic and provide some guidance in how to get started. You will collaborate in groups of 3-5 people to come up with a clever implementation that relates to the problem. Each group will present their progress and potentially some results on Friday afternoon.

Everyone (optional)

• 14/07/2023
• Center of Leuven

Beleuvenissen Festival

For those who are still in Leuven in the evening, there will be a free open-air music festival called “Beleuvenissen” at 4 locations in the centre! We can meet up to go to one of the locations together, such as the Oude Markt (NL Rock music) or the Vismarkt (Electro music).

More info on the event can be found here