| Title |
Radialinės bazės funkcijos neuroninio tinklo taikymas laiko eilučių, pagrįstų nereguliariu rekonstravimu, prognozavime / |
| Translation of Title |
Radial basis function neural network application to time series prediction, which are based on an non-regular embedding. |
| Authors |
Drūlytė, Miglė |
| Full Text |
|
| Pages |
70 |
| Keywords [eng] |
time series prediction ; attractor embedding ; time lags ; RBF ; neural network |
| Abstract [eng] |
The goal of this work is to show that the time series prediction using radial basis function neural network can be improved by using time series reconstruction into time delay space with non-uniform time delays. The time lags and embedding dimension are two factors that determine the attractor reconstruction with delay coordinate method. The embedding dimension describes the dimension of the time delay space. Identification of embedding parameters includes not only optimization of time lags but also determination of optimal dimension of the reconstructed phase space. The forecasting of reconstructed time series is based on RBF (radial basis function) neural network. The experiments in this work with time series shows that proposed method can significantly improve the prediction accuracy. |
| Type |
Master thesis |
| Language |
Lithuanian |
| Publication date |
2013 |
