Eine Plattform für die Wissenschaft: Bauingenieurwesen, Architektur und Urbanistik
Eine Plattform für die Wissenschaft: Bauingenieurwesen, Architektur und Urbanistik
Mackey–Glass Chaotic Time Series Forecasting by Using Self-Organizing Fuzzy Inference Network
In this paper, a new soft computing technique for creating self-organizing fuzzy inference network (SOFIN) which generates the network automatically from learning data by combining fuzzy c-means clustering method (FCM), learning from examples (LFE) technique and gradient descent method (GDM) is proposed in order to solve the problem of Mackey–Glass chaotic time series forecasting. First, the FCM is used to establish the centers of active functions in first layer of network. Second, the LFE technique is used to obtain the weights connecting the second layer to the third layer of the network which represents a fuzzy rule-base. After the structure of network is established, the last layer of the network is trained by using the GDM. The main contribution of this technique is to propose a time series forecasting model with lower complexity of computing and higher forecasting accuracy than the other typical forecasting techniques. This technique has been validated through the forecasting analysis of Mackey–Glass chaotic time series to be used as a benchmark data set for time series forecasting. It is reasonable to predict a non-stationary time series with long term such as electricity demand, stock prices and the spread of pandemic disease in particular.
Mackey–Glass Chaotic Time Series Forecasting by Using Self-Organizing Fuzzy Inference Network
In this paper, a new soft computing technique for creating self-organizing fuzzy inference network (SOFIN) which generates the network automatically from learning data by combining fuzzy c-means clustering method (FCM), learning from examples (LFE) technique and gradient descent method (GDM) is proposed in order to solve the problem of Mackey–Glass chaotic time series forecasting. First, the FCM is used to establish the centers of active functions in first layer of network. Second, the LFE technique is used to obtain the weights connecting the second layer to the third layer of the network which represents a fuzzy rule-base. After the structure of network is established, the last layer of the network is trained by using the GDM. The main contribution of this technique is to propose a time series forecasting model with lower complexity of computing and higher forecasting accuracy than the other typical forecasting techniques. This technique has been validated through the forecasting analysis of Mackey–Glass chaotic time series to be used as a benchmark data set for time series forecasting. It is reasonable to predict a non-stationary time series with long term such as electricity demand, stock prices and the spread of pandemic disease in particular.
Mackey–Glass Chaotic Time Series Forecasting by Using Self-Organizing Fuzzy Inference Network
J. Inst. Eng. India Ser. B
Choe, Myong-Song (Autor:in) / Ri, Kang-Song (Autor:in) / Ryang, Kyong-Il (Autor:in) / Choe, Un-Chol (Autor:in) / Ri, Chol-Guk (Autor:in)
Journal of The Institution of Engineers (India): Series B ; 104 ; 423-432
01.04.2023
10 pages
Aufsatz (Zeitschrift)
Elektronische Ressource
Englisch
Analytical Solution of Modified Mackey-Glass Equation
| Springer Verlag | 2022
Analytical Solution of Modified Mackey-Glass Equation
| TIBKAT | 2023
Wave Forecasting In Arabian Sea Using Adaptive Network Fuzzy Inference System
| HENRY – Bundesanstalt für Wasserbau (BAW) | 2010