This volume is a collection of contributions resulting from last year’s Workshop on Computational Optimization (WCO2012), which focused on algorithms based on metaheuristic methods applied to real-life problems.

Written by Angelova et al., the first paper addresses fermentation process models using intuitional fuzzy logic (IFL) to assess the performance of genetic algorithms (GAs). In particular, multipopulation GAs (MpGAs) are used for parameter identification of fermentation agents.

The second paper, by Rostami et al., considers recommender systems (RS) applied to item-based collaborative filtering (CF), based on user rating matrices. To overcome sparsity in item-based CF, the authors propose a new optimization approach based on graph representation of an item similarity matrix.

Deleplanque and Quilliot, in the third paper, deal with on-demand transportation (ODT) systems, addressing the dial-and-ride problem with split loads (DARPSL). The authors propose a general framework for the model, and describe two algorithms, one for the dial-and-ride problem (DARP) and one for DARPSL. They provide numerical experiments for both of these algorithms.

In the paper by Fidanova et al., ant colony optimization (ACO) and a GA are used for parameter identification in a model that simulates E. coli batch cultivation. The objective function was formulated as the distance between model-predicted outcomes (a system of nonlinear differential equations) and the experimental data. ACO and GA algorithms were compared for parameter optimization. Two distances are compared: the least-squares regression and the Hausdorff distance. The results show that an objective function with a Hausdorff distance metric produces better results for model and high-parameter accuracy, although it is more expensive computationally than the least-squares distance. This paper stands out for the clarity of its exposition on methods employed.

A paper by Akeb et al. solves the strip-packing problem (SPP) part of the cutting-and-packing (C&P) problem in operational research. This problem occurs when packing circular objects of different radii into a container. The authors propose an algorithm that combines a beam search with a restarting strategy and a look-ahead strategy.

Anghinolfi et al. propose using the ACO approach to solve the problem of how to optimally route the wiring in large-scale modular skins of robots. They address the problem statement as a graph representing a patch, modeling the skin wiring problem as a constrained spanning forest problem.

Quilliot et al. address the problem of multiprocessor scheduling with constraints that forbid interruption in the use of the processors, and recast it in the form of pyramidal shape functions that involve the time function T. They propose several minimizations of a makespan non-idle problem in polynomial-time algorithms.

Roeva and Slavov propose a hybrid firefly algorithm (FA) and a GA for a proportional-integral-derivative (PID) controller that controls feed rate and maintains glucose concentration at a desired set point for an E. coli fed-batch cultivation process.

In the final contribution, Stachurski considers a new formulation of a Broyden quasi-Newton restricted convex class involving oblique projections and their properties. Formal proof is provided for theoretical equivalence in both updating formulas. This research offers new possibilities in convergence analysis of quasi-Newton methods for minimization. It opens up the possibility of investigating convergence for problems with a singular Hessian matrix for the cost function. This paper stands out in terms of its contribution to quasi-Newton algorithms.

The volume is a collection of highly specialized problems addressing the real-life minimization of problems characterized by non-differentiable, discontinuous, noisy, or dynamical objective functions and constraints, as they occur in bioreactors, robot skin wiring design, strip packing, project scheduling, and tuning of PID controllers, which require metaheuristic algorithms to avoid huge computational resource requirements.