Now showing 1 - 10 of 16
  • Publication
    FINANCIAL FORECASTING USING NEURAL NETWORKS
    (2016-12-10)
    This paper presents an application of neural networks to financial time-series forecasting. No additional indicators, but only the information contained in the sales time series was used to model and forecast stock exchange index. The forecasting is carried out by two different neural network learning algorithms – error backpropagation and Kohonen self-organising maps. The results are presented and their comparative analysis is performed in this article.
  • Publication
    DATA PREPROCESSING METHODS FOR INTERVAL BASED NEURAL NETWORK PREDICTION
    The paper examines a task of forecasting stock prices of Riga Stock exchange by the use of interval value prediction approach, which is carried out by modified Kohonen neural network learning algorithm. The data preprocessing methods are analyzed and implemented here to solve stock prices prediction task. The proposed data preprocessing methods has been experimentally tested with two types of artificial neural networks.
  • Publication
    RETAIL TURNOVER PREDICTION USING MODULAR ARTIFICIAL NEURAL NETWORKS
    The paper focuses on the retail turnover prediction with artificial neural networks. The artificial neural networks have the potential to learn complex, non-linear relationships within data. The main disadvantage is that neural networks are “black boxes”, so the user cannot explain the obtained results and relationships between data. The modular neural networks allow obtaining more appropriate results by splitting the task into subtasks, thus giving the user more information in the output. In many cases an additional advantage of modular neural network is more precise prediction results, which will be shown in the experimental part of this paper.
    Scopus© Citations 1
  • Publication
    SORTING ALGORITHMS REALIZATION AND THEIR FEATURES
    (2022-05-07)
    Ivars Cabuļevs
    ;
    The performance of computers and programs is the most important thing for a user nowadays. Sorting algorithms appeared in the 19th century, but nowadays, developers often forget about the effectiveness of these algorithms and always use only a couple of algorithms, which are not always the best solution for certain tasks. This slows down the performance of certain important applications for professionals as well as for ordinary users. In this work was told and implemented different numerical sorting algorithms. The development environment here was “Microsoft Visual Studio 2017”. Was used the C++ programming language. Knowledge of sorting algorithms will always help you to optimize your program (if there used sorting), which will have a positive impact on user feedback about your application also this knowledge has a positive effect on the thought processes, allowing you to make the right decisions in the shortest time.
  • Publication
    EVOLUTIONARY ALGORITHMS LEARNING METHODS IN STUDENT EDUCATION
    Teaching experience shows that during educational process student perceive graphical information better than analytical relationships. As a possible solution, there could be the use of package Matlab in realization of different algorithms for IT studies. Students are very interested in modern data mining methods, such as artificial neural networks, fuzzy logic, clustering and evolution methods. Series of research were carried out in order to demonstrate the suitability of the Matlab for the purpose of visualization of various simulation models of some data mining disciplines – particularly genetic algorithms. Nowadays the possibilities of evolutionary algorithms are widely used in many optimization and classification tasks. There are four paradigms in the world of evolutionary algorithms: evolutionary programming, evolution strategies, genetic algorithms and genetic programming. This paper analyses present-day approaches of genetic algorithms and genetic programming and examines the possibilities of genetic programming that will be used in further research. Genetic algorithm learning methods are often undeservedly forgotten, although the implementation of their algorithms is relatively strong and can be implemented even for students. In the research part of the study the modelling capabilities in data mining studies were demonstrated based on genetic algorithms and real examples. We assume that students already have prior knowledge of genetic algorithms.
  • Publication
    APPROACHES AND SOLUTIONS FOR SIGN LANGUAGE RECOGNITION PROBLEM
    The goal of the paper is reviewing several aspects of Sign Language Recognition problems focusing on Artificial Neural Network approach. The lack of automated Latvian Sign Language has identified and proposals of how to develop such a system have made. Tha authors use analytical, statistical methods as well as practical experiments with neural network software. The main results of the paper are description of main Sign Language Recognition problem solving methods with Artificial Neural Networks and directions of future work based on authors’ previous expertise.
  • Publication
    LATVIAN SIGN LANGUAGE RECOGNITION CLASSIFICATION POSSIBILITIES
    There is a lack of automated sign language recognition system in Latvia while many other countries have been already equipped with such a system. Latvian deaf society requires support of such a system which would allow people with special needs to enhance their communication in governmental and public places. The aim of this paper is to recognize Latvian sign language alphabet using classification approach with artificial neural networks, which is a first step in developing integral system of Latvian Sign Language recognition. Communication in our daily life is generally vocal, but body language has its own significance. It has many areas of application like sign languages are used for various purposes and in case of people who are deaf and dumb, sign language plays an important role. Gestures are the very first form of communication. The paper presents Sign Language Recognition possibilities with centre of gravity method. So this area influenced us very much to carry on the further work related to hand gesture classification and sign’s clustering.
  • Publication
    The Influence of Hidden Neurons Factor on Neural Nework Training Quality Assurance
    The work shows the role of hidden neurons in the multilayer feed-forward neural networks. The numeric expression of hidden neurons is usually determined in each case empirically. The methodology for determining the number of hidden neurons are described. The neural network based approach is analyzed using a multilayer feed-forward network with backpropagation learning algorithm. We have presented neural network implementation possibility in bankruptcy prediction (the experiments have been performed in the Matlab environment). On the base of bankruptcy data analysis the effect of hidden neurons to specific neural network training quality is shown. The conformity of theoretical hidden neurons to practical solutions was carried out.
    Scopus© Citations 2
  • Publication
    DATA MINING TEACHING POSSIBILITIES USING MATLAB
    The teaching experience in the study process shows that students are better at perceiving graphical information rather than analytical relationships. Many training courses run on models that were previously available only in mathematics or physics. The use of Matlab package for the implementation of various algorithms in the Information Technology fields could be a possible solution. Often, the analytical solution is much simpler than the visual Matlab model, but for the purposes of perspective training it gives understanding of the usefulness of using such models. In the previous articles the authors had given examples of how Matlab's possibilities could be used for economic research purposes (optimal tax rate searching and modelling market equilibrium price). Students are very interested in modern data mining methods, such as artificial neural networks. In the research part of the study, the modelling capabilities in data mining studies are demonstrated by neural network examples.