Traditional means of studying environmental economics and management problems consist of optimal control and dynamic game models that are solved for optimal or equilibrium strategies. Notwithstanding the possibility of multiple equilibria, the models’ users—managers or planners—will usually be provided with a single optimal or equilibrium strategy no matter how reliable, or unreliable, the underlying models and their parameters are. In this paper we follow an alternative approach to policy making that is based on viability theory. It establishes “satisficing” (in the sense of Simon), or viable, policies that keep the dynamic system in a constraint set and are, generically, multiple and amenable to each manager’s own prioritisation. Moreover, they can depend on fewer parameters than the optimal or equilibrium strategies and hence be more robust. For the determination of these (viable) policies, computation of “viability kernels” is crucial. We introduce a MATLAB application, under the name of VIKAASA, which allows us to compute approximations to viability kernels. We discuss two algorithms implemented in VIKAASA. One approximates the viability kernel by the locus of state space positions for which solutions to an auxiliary cost-minimising optimal control problem can be found. The lack of any solution implies the infinite value function and indicates an evolution which leaves the constraint set in finite time, therefore defining the point from which the evolution originates as belonging to the kernel’s complement. The other algorithm accepts a point as viable if the system’s dynamics can be stabilised from this point. We comment on the pros and cons of each algorithm. We apply viability theory and the VIKAASA software to a problem of by-catch fisheries exploited by one or two fleets and provide rules concerning the proportion of fish biomass and the fishing effort that a sustainable fishery’s exploitation should follow.

We describe an optimization-based method for tackling the classic image processing problem known as edge detection and we formulate it in the form of a classification one. The novelty of the approach is in the use of spherical separation as a classification tool in the image processing framework. Spherical separation consists in separating by means of a sphere two given discrete point-sets in a finite dimensional Euclidean space; in our context the two sets are the edge points and the non-edge points, respectively, in the digital representation of a given image. Assuming that the center of the sphere is fixed, the problem reduces to the minimization of a convex and nonsmooth function of just one variable, which can be effectively solved by means of an “ad hoc” bisection method. The results of our experiments on some edge detection benchmark images are provided.

We investigate constrained first order techniques for training support vector machines (SVM) for online classification tasks. The methods exploit the structure of the SVM training problem and combine ideas of incremental gradient technique, gradient acceleration and successive simple calculations of Lagrange multipliers. Both primal and dual formulations are studied and compared. Experiments show that the constrained incremental algorithms working in the dual space achieve the best trade-off between prediction accuracy and training time. We perform comparisons with an unconstrained large scale learning algorithm (Pegasos stochastic gradient) to emphasize that our choice can remain competitive for large scale learning due to the very special structure of the training problem.

This paper presents a new heuristic for generating scenarios for two-stage stochastic programs. The method uses copulas to describe the dependence between the marginal distributions, instead of the more common correlations. The heuristic is then tested on a simple portfolio-selection model, and compared to two other scenario-generation methods.

Convex optimization methods are used for many machine learning models such as support vector machine. However, the requirement of a convex formulation can place limitations on machine learning models. In recent years, a number of machine learning methods not requiring convexity have emerged. In this paper, we study non-convex optimization problems on the Stiefel manifold in which the feasible set consists of a set of rectangular matrices with orthonormal column vectors. We present examples of non-convex optimization problems in machine learning and apply three nonlinear optimization methods for finding a local optimal solution; geometric gradient descent method, augmented Lagrangian method of multipliers, and alternating direction method of multipliers. Although the geometric gradient method is often used to solve non-convex optimization problems on the Stiefel manifold, we show that the alternating direction method of multipliers generally produces higher quality numerical solutions within a reasonable computation time.

Using a dynamic network formation model, solved numerically, we study banks’ rollover decisions. We find that when the existence of linkages between market participants generates an informational externality, the newly formed network is conditioned by past architectures. Moreover, this inertia is strongly dependent on macroeconomic conditions, such as investors’ risk appetite. Simulations show that for intermediate values of the risk appetite’s parameter the financial network exhibits tipping points, i.e., the inability to maintain a threshold number of linkages may push the market into a gridlock. In this context, we study also how policy instruments, such as taxes and subsidies, affects debt rollover. Since a reduction in the policy level plays the same role as an improvement in economic fundamentals, the creation of interbank connections can be stimulated by it. Thus, in order to restart lending after a major stress situation in the interbank market a considerable reduction in the policy level is required, advising a counter-cyclical policy similar to the ones recently proposed with respect to capital requirements.

This paper presents a new algorithm for learning the structure of a special type of Bayesian network. The conditional phase-type (C-Ph) distribution is a Bayesian network that models the probabilistic causal relationships between a skewed continuous variable, modelled by the Coxian phase-type distribution, a special type of Markov model, and a set of interacting discrete variables. The algorithm takes a data set as input and produces the structure, parameters and graphical representations of the fit of the C-Ph distribution as output. The algorithm, which uses a greedy-search technique and has been implemented in MATLAB, is evaluated using a simulated data set consisting of 20,000 cases. The results show that the original C-Ph distribution is recaptured and the fit of the network to the data is discussed.

We present the parallelization of a linear programming solver using a primal-dual interior point method on one of the heterogeneous processors, namely the Cell BE processor. Focus is given to Cholesky factorization as it is the most computationally expensive kernel in interior point methods. To make it easier to develop and port to other heterogeneous systems, we propose a two-phase implementation procedure where we first develop a shared-memory multithreaded application that executes only on the main processor, and then offload the compute-intensive tasks to execute on the synergistic processors (Cell accelerator cores). We used parent–child supernode amalgamation to increase sizes of the blocks, but we noticed that the existence of many small blocks cause significant performance degradation. To reduce the overhead of small blocks, we extend the block fan-out algorithm such that small blocks are aggregated into large blocks without adding extra zeros. We also use another type of amalgamation that can merge any two consecutive supernodes and use it to avoid having very small blocks in a composed block. The suggested block aggregation method is able to speedup the whole LP solver of up to 2.5 when compared to using parent–child supernode amalgamation alone.

Blood service operations are a key component of the healthcare system all over the world and yet the modeling and analysis of such systems from a supply chain network optimization perspective has been lacking due to their associated unique challenges. In this paper, we develop a network optimization model for the complex supply chain of human blood, which is a life-saving, perishable product. In particular, we consider a regionalized blood banking system consisting of collection sites, testing and processing facilities, storage facilities, distribution centers, as well as points of demand, which, typically, include hospitals. Our multicriteria system-optimization approach on generalized networks with arc multipliers captures many of the critical issues associated with blood supply chains such as the determination of the optimal allocations, and the induced supply-side risk, as well as the induced cost of discarding the waste, while satisfying the uncertain demands as closely as possible. The framework that we present is also applicable, with appropriate modifications, to the optimization of other supply chains of perishable products.

Stochastics affects the optimal design of a network. This paper examines the single-source single-commodity stochastic network design problem. We characterize the optimal designs under demand uncertainty and compare with the deterministic counterparts to outline the basic structural differences. We do this partly as a basis for developing better algorithms than are available today, partly to simply understand what constitutes robust network designs.