Browsing by Author "Савка, Н. Я."

Filter results by typing the first few letters
Now showing 1 - 1 of 1
  • Thumbnail Image
    Item
    Моделювання податково-боргової складової фінансової безпеки на основі штучних нейронних мереж
    (Видавництво Львівської політехніки, 2015) Михальчук, Н. М.; Савка, Н. Я.
    Запропоновано методику моделювання податково-боргової компоненти фінансової безпеки держави. Теоретично обґрунтовано доцільність поєднання в методиці нормативно-індикативної оцінки з нейромережевими технологіями. Проаналізовано вплив податкової заборгованості на фінансову безпеку за двома індикаторами: податковим навантаженням та рівнем тінізації економіки. Topicality of the problem of restructuring basic approaches to principles, forms, methods and tools for providing financial security of the state is justified in the article. The current concept, which is legally approved, is based on regulatory and indicative analysis and has the nature of ascertaining of current trends; however it doesn’t allow analyzing the future transformations. The authors prove complexity of applying classical evaluation methods from both mathematic and logical point of view through the features of political and economic situation in Ukraine and through nonlinearity relation between the events and the processes within security bounds. Accordingly, the authors argue advisability and preferences in the formation of new, more progressive methods for diagnosing the state of financial security at the macro level, especially by the tax-debt component. In the process of research there was developed methodology for modeling tax-debt component of state financial security, which was developed by using combination of the normative and indicative evaluation with the neural network technologies. Efficiency of methodology was verified by using analysis of impact of the tax arrears on the financial security by using two indicators: tax load and the level of the shadow economy. As the result of expert evaluations there was defined a range of limit values for selected indicators and it was found that their values show crisis trends. The model, which was developed based on the artificial neural networks with the radial basis functions, demonstrates interconnection between the level of the tax debt and the level of the shadow economy and refutes the dependence between arrears and tax burden. Those results are adequate to the current economic situation.