• Publications
  • Influence
Neural network approach to forecasting of quasiperiodic financial time series
TLDR
A novel neural network approach to forecasting of financial time series based on the presentation of the series as a combination of quasiperiodic components is presented. Expand
  • 87
  • 3
Multilayer Neuro-fuzzy Network for Short Term Electric Load Forecasting
TLDR
A novel neuro-fuzzy network architecture and learning algorithms are proposed, which enable high-rate processing of information given in different measurements scales (quantitative, ordinal, and nominal). Expand
  • 21
An Optimal Algorithm for Combining Multivariate Forecasts in Hybrid Systems
TLDR
The problem of combining of multivariate forecasts produced by different components in a hybrid system is considered. Expand
  • 2
Robust Learning Algorithm for Networks of Neuro-Fuzzy Units
TLDR
A new learning algorithm based on a robust criterion is proposed that allows effective handling of outliers. Expand
  • 4
Nonlinear Visualization of Incomplete Data Sets
TLDR
We propose a modified learning procedure for the autoassociative neural network that directly takes into account missing values in the original data set. Expand
  • 3
Feedforward neural network with a specialized architecture for estimation of the temperature influence on the electric load
TLDR
A novel architecture of a feedforward neural network is proposed that provides separation of temperature influence from other factors and its analysis in an explicit form. Expand
  • 4
Neuro-fuzzy Elman network for short-term electric load forecasting
TLDR
A modified architecture of Elman-type recurrent neural network is proposed. Expand
  • 1
  • PDF
Harmonic Components Detection in Stochastic Sequences Using Artificial Neural Networks
The problem of hidden harmonic components detection in stochastic sequences occurs quite often in practice and, first of all, in technical diagnostics and monitoring of signals of various natures. InExpand
Function Decomposition Network
TLDR
Novel neural network architecture is proposed to solve the nonlinear function decomposition problem. Expand
...
1
2
...