Abstract | ||
---|---|---|
Most traditional time series models are based on local methods (in time), which means assuming that the time series can be fully and locally (in time) characterized with a finite embedding space. There are many situations in which simple regression can not help find the temporal structural in time series. In this research, a Markovian architecture: Markov Gated Experts, has been developed based on nonlinearly gated experts. This paper discusses the statistical framework and compares the performance of Markov gated experts to gated experts on both computer generated time series and real world data. Compared with the original method, Markov gated experts are more powerful in finding the underlying temporal structure, and are therefore a more powerful analytical and forecasting model for non-stationary and structurally changing time series. |
Year | DOI | Venue |
---|---|---|
1997 | 10.1109/ICNN.1997.614215 | 1997 IEEE INTERNATIONAL CONFERENCE ON NEURAL NETWORKS, VOLS 1-4 |
Keywords | Field | DocType |
time series model,information systems,neural networks,computer architecture,feedforward neural networks,computer science,time series,predictive models,regression,structural change,time series analysis,markov processes,neural nets,generation time | Time series,Embedding,Markov process,Regression,Computer science,Markov chain,Artificial intelligence,Simple linear regression,Artificial neural network,Machine learning | Conference |
Citations | PageRank | References |
1 | 0.39 | 2 |
Authors | ||
2 |
Name | Order | Citations | PageRank |
---|---|---|---|
shanming shi | 1 | 1 | 0.39 |
Andreas S. Weigend | 2 | 576 | 112.30 |