T. Koskela, M. Varsta, J. Heikkonen, K. Kaski, "Recurrent SOM with Local Linear Models in Time Series Prediction", in Proc. of ESANN'98, Bruges (Belgium), D-Facto pub. (Brussels), pp. 167-172, 1998. Home/Search Document Details and Download
Summary Related Articles Check |
This paper is cited in the following contexts:
A General Framework for Self-Organizing Structure.. - Hammer, Micheli.. (Correct)
....approaches alternatively enlarge SOM by recurrent dynamics such as leaky integrators or more general recurrent connections which allow the recursive processing of sequences. Examples are the temporal Kohonen map (TKM) 7] the recursive SOM (RecSOM) 45, 46, 47] or the approaches proposed in [9, 24, 25, 31]. The SOM for structured data (SOMSD) 17, 18, 41] constitutes a recursive mechanism capable of processing tree structured data and thus also sequences in an unsupervised way. Alternative models for unsupervised time series processing use for example hierarchical network architectures. An overview ....
....internally represented by the neural map. An appropriate choice of the form of internal representations allows to recover the respective mechanism. Moreover, the dynamic of supervised recurrent and recursive networks can be integrated in the general framework as well. The approaches reported in [9, 24, 31] can be simulated with slight variations of parts of the framework. Hence we obtain a uniform formulation which allows a straightforward investigation of possible learning algorithms and theoretical properties of several important approaches proposed in the literature for SOMs with recurrence. ....
[Article contains additional citation context not shown here]
T. Koskela, M. Varsta, J. Heikkonen, and K. Kaski. Recurrent SOM with local linear models in time series prediction. In M.Verleysen (ed.), 6th European Symposium on Artificial Neural Networks, pages 167--172, De facto, 1998.
Applying LSTM to Time Series Predictable Through.. - Gers, Eck, Schmidhuber (2001) (2 citations) (Correct)
....100 points. We run tests for stepwise prediction and fully iterated prediction, where the output is clamped to the input for 100 steps. For the experiments with MLPs the setup was as described for the MG data but with an input embedding of the last 9 time steps as in Koskela, Varsta and Heikkonen [21]. FIR laser Previous Work Results are listed in Table 2. Linear prediction is no better than predicting the data mean. Wan [22] achieved the best results submitted to the original Santa Fe contest. He used a Finite Input Response Network (FIRN) 25 inputs and 12 hidden units) a method similar to ....
....the best results submitted to the original Santa Fe contest. He used a Finite Input Response Network (FIRN) 25 inputs and 12 hidden units) a method similar to a TDNN. Wan improved performance by replacing the last 25 predicted points by smoothed values (sFIRN) Koskela, Varsta and Heikkonen [21] compared recurrent SOMs (RSOMs) and MLPs (trained with the Levenberg Marquardt algorithm) with an input embedding of dimension 9 (an input window with the last 9 values) Bakker et al. 10] used a mixture of predictions and true values as input (Error Propagation, EP) Then Principal Component ....
[Article contains additional citation context not shown here]
T. Koskela, M. Varsta, J. Heikkonen, and K. Kaski, \Recurrent SOM with local linear models in time series prediction," in 6th European Symposium on Articial Neural Networks. ESANN'98. Proceedings. D-Facto, Brussels, Belgium, pp. 167{ 72, 1998.
Applying LSTM to Time Series Predictable Through.. - Gers, Eck, Schmidhuber (2001) (2 citations) (Correct)
....the next 100 points (Figure 5) We run tests for stepwise prediction and fully iterated prediction, where the output is clamped to the input for 100 steps. For the experiments with MLPs the setup was as described for the Mackey Glass data but with an input embedding of the last 9 time steps as in Koskela, Varsta, Heikkonen, and Kaski (1998). 4.1 Previous Work Results are listed in Table 4.1. Linear prediction is no better than predicting the data mean. Wan (1994) achieved the best results submitted to the original Santa Fe contest. He used a Finite Input Response Network (FIRN) 25 inputs and 12 hidden units) a method similar to ....
....t. Middle: Cell states s c . Bottom: Activations of the gates. Technical Report No. IDSIA IDSIA 22 00 10 0 100 200 0 200 400 600 800 test time 1000 trainig 0 0 20 40 60 80 time 100 test 200 100 Figure 5: FIR laser Data (Set A) from the Santa Fe time series prediction competition. Koskela, Varsta, Heikkonen, and Kaski (1998) compared recurrent SOMs (RSOMs) and MLPs (trained with the Levenberg Marquardt algorithm) with an input embedding of dimension 9 (an input window with the last 9 values) Bakker, Schouten, Giles, Takens, and Bleek (2000) used a mixture of predictions and true values as input (Error Propagation, ....
[Article contains additional citation context not shown here]
Koskela, T., Varsta, M., Heikkonen, J., & Kaski, K. (1998). Recurrent SOM with local linear models in time series prediction. In 6th european symposium on articial neural networks. esann'98. proceedings. d-facto, brussels, belgium (pp. 167-72).
Bibliography of Self-Organizing Map (SOM) Papers.. - Merja Oja, Samuel.. (2002) Self-citation (Kaski) (Correct)
No context found.
Koskela, T., Varsta, M., Heikkonen, J., and Kaski, K. (1998). Recurrent SOM with local linear models in time series prediction. In 6th European Symposium on Artificial Neural Networks. ESANN'98. Proceedings, pages 167--72, Brussels, Belgium. D-Facto.
Double Quantization Forecasting Method for Filling Missing.. - Geoffroy Simon John (2004) (Correct)
No context found.
T. Koskela, M. Varsta, J. Heikkonen, K. Kaski, "Recurrent SOM with Local Linear Models in Time Series Prediction", in Proc. of ESANN'98, Bruges (Belgium), D-Facto pub. (Brussels), pp. 167-172, 1998.
Self-Organizing Neural Networks for Sequence Processing - Strickert (Correct)
No context found.
T. Koskela, M. Varsta, J. Heikkonen, and K. Kaski. Recurrent SOM with local linear models in time series prediction. In M. Verleysen, editor, European Symposium on Artificial Neural Networks (ESANN), pages 167--172. D-facto Publications, 1998.
Long-Term Time Series Forecasting Using Self-Organizing.. - Simon, Lendasse, al. (2003) (Correct)
No context found.
T. Koskela, M. Varsta, J. Heikkonen, and K. Kaski, "Recurrent SOM with Local Linear Models in Time Series Prediction", Proc. of ESANN'98, 6th European Symposium on Artificial Neural Networks, D-Facto, Brussels, Belgium, pp. 167-172, April 1998.
A General Framework for Unsupervised Processing of.. - Hammer, Micheli, al. (2002) (1 citation) (Correct)
No context found.
T. Koskela, M. Varsta, J. Heikkonen, and K. Kaski. Recurrent SOM with local linear models in time series prediction. In M.Verleysen (ed.), 6th European Symposium on Artificial Neural Networks, pages 167--172, De facto, 1998.
Mathematical Aspects of Neural Networks - Hammer (2003) (Correct)
No context found.
T. Koskela, M. Varsta, J. Heikkonen, and K. Kaski. Recurrent SOM with local linear models in time series prediction. In M.Verleysen, editor, 6th European Symposium on Artificial Neural Networks,pages 167--172, De facto, 1998. 200
Double SOM for long-term time series prediction - Geoffroy Simon Amaury (2003) (Correct)
No context found.
T. Koskela, M. Varsta, J. Heikkonen, and K. Kaski, "Recurrent SOM with Local Linear Models in Time Series Prediction", in Proc. of ESANN'98, pp. 167-172, D-Facto, Brussels, 1998.
Online articles have much greater impact More about CiteSeer.IST Add search form to your site Submit documents Feedback
CiteSeer.IST - Copyright Penn State and NEC