Abstract

In this paper we proposed two new strategies for initialization and training of Radial Basis Function (RBF) Neural Network. The first approach takes into consideration the "error" between the input vector p of the network and the x-axis, which are the centers of radial functions. The second approach takes into account the "error" between the input vector p and the network output. In order to check the performances of these strategies, we used Brazilian financial market data for the RBF networks training, specifically the adjusted prices of the 10 greater weighted shares in the Bovespa index at the time of data collection - from April 8th , 2009 to October 31th , 2014. The first approach presented a 52% of improvement in the mean squared error (MSE) compared to the standard RBF network, while the improvement for the second approach was 38%. The strategies proved to be consistent for the time series tested, in addition to having a low computational cost. It is proposed that these strategies be improved by testing them with the Levenberg-Marquardt algorithm.

Abstract

The topic of the study is relevant due to the fact that abrupt changes in the economy and in the financial system, particularly those caused by the coronavirus pandemic, have changed all existing development forecasts. At the same time, a whole range of new risks for the financial system appeared. In this regard, the development of a forecast for the development of the financial sector in various scenarios, taking into account risks and opportunities, seems to be the most important task for the Russian economy. The purpose of the study is to develop a medium-term forecast for the development of the financial sector in the Russian Federation. The subject of this research is the dynamics of the development of the financial system of the Russian Federation in the medium term. The research method or methodology is based on the study of key current and retrospective data on the state of the financial sector of the Russian Federation, highlighting the most important risks for the financial sector and building a medium-term forecast. As a result of the work carried out, the key components of the budgetary sphere, financial and credit markets have been analyzed, and the impact of the deterioration in the external environment has been assessed. Thus, the work presents the assessments of the prerequisites for the forecast and developed three possible scenarios for the development of the financial sector in Russia. The scientific novelty of the research lies in producing unique forecasts for the development of the financial sector in Russia. The recommendations obtained as a result of the study indicate the need to develop mechanisms for responding to certain risks for the financial system, taking into account the possible consequences of such an impact presented in the forecast. The project was carried out as part of research work by the Laboratory for Structural Research of the IAER at the Russian Presidential Academy of National Economy and Public Administration in 2021.