Informatica

This work compares different algorithms to replace the genetic optimizer used in a recent methodology for creating realistic and computationally efficient neuron models. That method focuses on single-neuron processing and has been applied to cerebellar granule cells. It relies on the adaptive-exponential integrate-and-fire (AdEx) model, which must be adjusted with experimental data. The alternatives considered are: i) a memetic extension of the original genetic method, ii) Differential Evolution, iii) Teaching-Learning-Based Optimization, and iv) a local optimizer within a multi-start procedure. All of them ultimately outperform the original method, and the last two do it in all the scenarios considered.

In this work, we perform an extensive theoretical and experimental analysis of the characteristics of five of the most prominent algebraic modelling languages (AMPL, AIMMS, GAMS, JuMP, and Pyomo) and modelling systems supporting them. In our theoretical comparison, we evaluate how the reviewed modern algebraic modelling languages match the current requirements. In the experimental analysis, we use a purpose-built test model library to perform extensive benchmarks. We provide insights on which algebraic modelling languages performed the best and the features that we deem essential in the current mathematical optimization landscape. Finally, we highlight possible future research directions for this work.

Machine Translation has become an important tool in overcoming the language barrier. The quality of translations depends on the languages and used methods. The research presented in this paper is based on well-known standard methods for Statistical Machine Translation that are advanced by a newly proposed approach for optimizing the weights of translation system components. Better weights of system components improve the translation quality. In most cases, machine translation systems translate to/from English and, in our research, English is paired with a Slavic language, Slovenian. In our experiment, we built two Statistical Machine Translation systems for the Slovenian-English language pair of the Acquis Communautaire corpus. Both systems were optimized using self-adaptive Differential Evolution and compared to the other related optimization methods. The results show improvement in the translation quality, and are comparable to the other related methods.

The heliostat field of Solar Central Receiver Systems takes up to 50% of the initial investment and can cause up to 40% of energetic loss in operation. Hence, it must be carefully optimized. Design procedures usually rely on particular heliostat distribution models. In this work, optimization of the promising biomimetic distribution model is studied. Two stochastic population-based optimizers are applied to maximize the optical efficiency of fields: a genetic algorithm, micraGA, and a memetic one, UEGO. As far as the authors know, they have not been previously applied to this problem. However, they could be a good option according to their structure. Additionally, a Brute-Force Grid is used to estimate the global optimum and a Pure-Random Search is applied as a baseline reference. Our empirical results show that many different configurations of the distribution model lead to very similar solutions. Although micraGA exhibits poor performance, UEGO achieves the best results in a reduced time and seems appropriate for the problem at hand.

This paper reviews the interplay between global optimization and probability models, concentrating on a class of deterministic optimization algorithms that are motivated by probability models for the objective function. Some complexity results are described for the univariate and multivariate cases.

The optimal financial investment (Portfolio) problem was investigated by leading financial organizations and scientists. Nobel prizes were awarded for the Modern Portfolio Theory (MPT) and further developments. The aim of these works was to define the optimal diversification of the assets depending on the acceptable risk level.

Relational mathematics, as it is studied in fields like mathematical economics and social choice theory for some time, provides a rich and general framework and appears to be a natural and direct way to paraphrase optimization goals, to represent user preferences, to justify fairness criterions, to cope with QoS or to valuate utility. Here, we will focus on the specific application aspects of formal relations in network design and control problems and provide the general concept of relational optimization. In relational optimization, we represent the optimization problem by a formal relation, and the solution by the set of maximal (or non-dominated) elements of this relation. This appears to be a natural extension of standard optimization, and covers other notions of optimality as well. Along with this, we will provide a set of fairness relations that can serve as maximizing relations in relational optimization according to various application needs, and we specify a meta-heuristic approach derived from evolutionary multi-objective optimization algorithms to approximate their maximum sets.

A simple Stock Exchange Game Model (SEGM) was introduced in Mockus (2003), to simulate the behavior of several stockholders using fixed buying-selling margins at fixed bank yield.

The concentration of a substrate in a solution can be measured using amperometric signals of biosensors: in fact the maximum (steady state) current is measured which is calibrated in the units of concentration. Such a simple method is not applicable in the case of several substrates. In the present paper, the problem of evaluation of concentrations of several substrates is tackled by minimizing the discrepancy between the observed and modeled transition processes of the amperometric signal.