The support system for the investor in currency market

This dissertation examines investment opportunities in the currency market, and analyses the theoretical and practical aspects of forecasting under uncertainty. At one time the world was formed of two distinct research areas: universal artificial intelligence theory and the theory of investment. The first area was influenced by the possibility of universal theory predictions for the emergence of what has been created for the results of various artificial intelligence algorithms and their systems. The second research area evolved as a rational prediction of the theory which laid the foundations for the emergence of modern portfolio theory. The thesis attempts to link these two scientific areas to the forecasting of currency market.

The main goal of this thesis is to create an investment decision-making support system targeting investors in the currency market, by adapting artificial intelligence algorithms and modern portfolio theory. In order to achieve the goal the following specific objectives are used: to create a forecasting model based on artificial intelligence algorithms, to integrate investment optimisation principles in predictive models, and to empirically substantiate the e‑ciency and reliability of the support system for investment in the currency market. The support system for investors was created with a targeted application of artificial intelligence algorithms to financial market forecasting, and by integrating them into modern portfolio theory.

The dissertation consists of an introduction, three chapters, a general conclusions, the reference list and a list of the author’s publications. The introductory chapter discusses the research problem and the relevance of the research, its aim and objectives, the research methodology, scientific innovation, practical significance and defended statements. The first chapter describes peculiarities of the financial markets, analyses of financial processes in the global economy, forecasting by artificial intelligence algorithms and an analysis of modern portfolio theory and investment strategies. The second chapter describes theoretical assumptions of the creation of artificial intelligence, the theoretical basis for adaptation of the Evolino recurrent neural networks (RNN) to productive work and provides an evaluation of portfolio e‑ciency. The third section presents the support system for investor in exchange market.

Four articles have been published on the dissertation topic: two in ISIWeb of Science journals and two in other reviewed journals. Nine papers were presented at international conferences: two as conference materials in ISI Proceedings, seven as conference materials in international conference proceedings.

Read electronic version of the book:

DOI: https://doi.org/10.20334/2310-M