Phase Transitions In Combinatorial Optimization Problems

A concise, comprehensive introduction to the topic of statistical physics of combinatorial optimization, bringing together theoretical concepts and algorithms from computer science with analytical methods from physics. The result bridges the gap between statistical physics and combinatorial optimization, investigating problems taken from theoretical computing, such as the vertex-cover problem, with the concepts and methods of theoretical physics. The authors cover rapid developments and analytical methods that are both extremely complex and spread by word-of-mouth, providing all the necessary basics in required detail. Throughout, the algorithms are shown with examples and calculations, while the proofs are given in a way suitable for graduate students, post-docs, and researchers. Ideal for newcomers to this young, multidisciplinary field.

The past few years have witnessed a substantial growth in the number of applications for optimization algorithms in solving problems in the field of physics. Examples include determining the structure of molecules, estimating the parameters of interacting galaxies, the ground states of electronic quantum systems, the behavior of disordered magnetic materials, and phase transitions in combinatorial optimization problems. This book serves as an introduction to the field, while also presenting a complete overview of modern algorithms. The authors begin with the relevant foundations from computer science, graph theory and statistical physics, before moving on to thoroughly explain algorithms - backed by illustrative examples. They include pertinent mathematical transformations, which in turn are used to make the physical problems tractable with methods from combinatorial optimization. Throughout, a number of interesting results are shown for all physical examples. The final chapter provides numerous practical hints on software development, testing programs, and evaluating the results of computer experiments.

Extremal Optimization: Fundamentals, Algorithms, and Applications introduces state-of-the-art extremal optimization (EO) and modified EO (MEO) solutions from fundamentals, methodologies, and algorithms to applications based on numerous classic publications and the authors’ recent original research results. It promotes the movement of EO from academic study to practical applications. The book covers four aspects, beginning with a general review of real-world optimization problems and popular solutions with a focus on computational complexity, such as "NP-hard" and the "phase transitions" occurring on the search landscape. Next, it introduces computational extremal dynamics and its applications in EO from principles, mechanisms, and algorithms to the experiments on some benchmark problems such as TSP, spin glass, Max-SAT (maximum satisfiability), and graph partition. It then presents studies on the fundamental features of search dynamics and mechanisms in EO with a focus on self-organized optimization, evolutionary probability distribution, and structure features (e.g., backbones), which are based on the authors’ recent research results. Finally, it discusses applications of EO and MEO in multiobjective optimization, systems modeling, intelligent control, and production scheduling. The authors present the advanced features of EO in solving NP-hard problems through problem formulation, algorithms, and simulation studies on popular benchmarks and industrial applications. They also focus on the development of MEO and its applications. This book can be used as a reference for graduate students, research developers, and practical engineers who work on developing optimization solutions for those complex systems with hardness that cannot be solved with mathematical optimization or other computational intelligence, such as evolutionary computations.

The authors stress the relative simplicity, efficiency, flexibility of use, and suitability of various approaches used to solve difficult optimization problems. The authors are experienced, interdisciplinary lecturers and researchers and in their explanations they demonstrate many shared foundational concepts among the key methodologies. This textbook is a suitable introduction for undergraduate and graduate students, researchers, and professionals in computer science, engineering, and logistics.

Phase transitions typically occur in combinatorial computational problems and have important consequences, especially with the current spread of statistical relational learning as well as sequence learning methodologies. In Phase Transitions in Machine Learning the authors begin by describing in detail this phenomenon, and the extensive experimental investigation that supports its presence. They then turn their attention to the possible implications and explore appropriate methods for tackling them. Weaving together fundamental aspects of computer science, statistical physics and machine learning, the book provides sufficient mathematics and physics background to make the subject intelligible to researchers in AI and other computer science communities. Open research issues are also discussed, suggesting promising directions for future research.

Computer science and physics have been closely linked since the birth of modern computing. In recent years, an interdisciplinary area has blossomed at the junction of these fields, connecting insights from statistical physics with basic computational challenges. Researchers have successfully applied techniques from the study of phase transitions to analyze NP-complete problems such as satisfiability and graph coloring. This is leading to a new understanding of the structure of these problems, and of how algorithms perform on them. Computational Complexity and Statistical Physics will serve as a standard reference and pedagogical aid to statistical physics methods in computer science, with a particular focus on phase transitions in combinatorial problems. Addressed to a broad range of readers, the book includes substantial background material along with current research by leading computer scientists, mathematicians, and physicists. It will prepare students and researchers from all of these fields to contribute to this exciting area.

This book constitutes the refereed proceedings of the 17th European Conference on Evolutionary Computation in Combinatorial Optimization, EvoCOP 2017, held in Amsterdam, The Netherlands, in April 2017, co-located with the Evo*2017 events EuroGP, EvoMUSART and EvoApplications. The 16 revised full papers presented were carefully reviewed and selected from 39 submissions. The papers cover both empirical and theoretical studies on a wide range of academic and real-world applications. The methods include evolutionary and memetic algorithms, large neighborhood search, estimation of distribution algorithms, beam search, ant colony optimization, hyper-heuristics and matheuristics. Applications include both traditional domains, such as knapsack problem, vehicle routing, scheduling problems and SAT; and newer domains such as the traveling thief problem, location planning for car-sharing systems and spacecraft trajectory optimization. Papers also study important concepts such as pseudo-backbones, phase transitions in local optima networks, and the analysis of operators. This wide range of topics makes the EvoCOP proceedings an important source for current research trends in combinatorial optimization.

Many physicists are not aware of the fact that they can solve their problems by applying optimization algorithms. Since the number of such algorithms is steadily increasing, many new algorithms have not been presented comprehensively until now. This presentation of recently developed algorithms applied in physics, including demonstrations of how they work and related results, aims to encourage their application, and as such the algorithms selected cover concepts and methods from statistical physics to optimization problems emerging in theoretical computer science.