TAAC'2011, Section Three: Applied mathematics

In this paper, I present an application for car color recognition. Images used as the inputs for the application were taken in a real world environment, so aspects, like light or weather conditions cause problems in this task. My approach is a simplified version of the current advanced approaches. This paper will illustrate the results that are to be expected with such an approach. I have defined parts of the car body where I picked color samples. Based on probability of each color presence in the sampled area, according to sophisticated rules, a dominant color is recognized.

On stage of budgeting bank treasury should evaluate the main targets of positions in major currencies, taking into account that exchange rate risk does not cause significant losses or even bankruptcy. Setting targets in its essence is a problem of portfolio optimization, where the contribution asset is the size measures the position of the currency; assets, from which there is a choice, are the currency and prices assets — exchange rates. Thus, to determine the optimal size of positions you can use basic analytical method, involving joint normal distribution of logarithmic returns series of exchange, or use other ways to allow more flexible modeling of multivariate distribution. Such tools available, particularly through the application of the theory of copulas. Therefore, the purpose of research is to study the effectiveness of the approach to problem solving that uses semi parametric copula method to recovery distribution.

The purpose of this work is developing a methodology for quantifying and analyzing social processes. We propose a mechanism for measuring the quantities which correspond to types of political activity, arguing the choice of analysis methods. As a result, practical studies have found links between dependence and species of political activity. The development of the proposed technique will allow to simulate the political situation in the state, to predict the possible consequences of policy decisions.

The paper discusses the application of Computational Intelligence in the study of complex economic system — the software market in Ukraine. There are identified factors influencing the market, scaling applied for their measurement, built descriptor market model as an abstract finite graph of factors relationships, constructed and trained neural network with the optimal architecture, selected by genetic algorithm, which reflects the influence of factors on the change in demand as a market indicator.

In this paper we consider the one-stage service model of determine objects flow by stationary servicing processor. We formulate a problem of generating optimal servicing strategies subject to two criteria. It is shown what if we take into account practically significant limitation on servicing discipline the problem become a non NP-hard and we can build a polynomial algorithm for generating effective servicing strategies. We introduce this algorithm and give the results of its approbation.

Work dedicated to the modeling process of budgeting at the production plant. It was a model that includes all elements of the process of budgeting is flexible and can be used practically in the business, increasing efficiency of its functioning. In addition, the model facilitates a better performance of such functions as budgeting planning process of any company, calculation of different variants of budgeting, monitoring and analysis of budgets.

Rationing problem for the case of fuzzy input data is explored. Its solution in accordance with the basic methods: proportional, Uniform Gains and Uniform losses is described. α-staged lexicographical order on a fuzzy set of vectors is introduced. Set of non-dominated alternatives is constructed and the problem is reduced to a 2-criteria decision making problem.

We consider the project of information-analytical system for accounting and monitoring of technical state of facilities (roads and engineering structures), which should help ease the preparation of various reporting forms and management decisions.

There were considered the main features of international trade of electricity between the Europecountries . The market was analized as a multi-agent system. I described the expediency of using the apparatus of game theory to assess scenarios of the system. This approach will allow to find the best import-export policy of Ukraine towards electricity trade, and also show the benefits of such relationship for Europe as an integrated economic system and solve the optimal ratio of generation, exports, imports, consumption of electricity.