Topology Optimization and Training of Recurrent Neural Networks with Pareto-Based Multi-objective Algorithms: A Experimental Study
The simultaneous topology optimization and training of neu- ral networks is a problem widely studied in the last years, specially for feedforward models. In the case of recurrent neural networks, the existing proposals attempt to only optimize the number of hidden units, since the problem of topology optimization is more difficult due to the feedback connections in the network structure. In this work, we make a study of the effects and difficulties for the optimization of network connections, hidden neurons and network training for dynamical recurrent models. In the experimental section , the proposal is tested in time series prediction problems.
KeywordsRecurrent Neural Networks Multi-Objective optimization
Unable to display preview. Download preview PDF.
- 5.Mandic, D.P., Chambers, J.: Recurrent Neural Networks for Prediction. John Wiley and sons, Chichester (2001)Google Scholar
- 6.Teo, J., Abbass, H.A.: Information-theoretic landscape analysis of neuro-controlled embodied organisms. Neural Computing and Applications 31(1), 80–89 (2004)Google Scholar
- 8.Cuéllar, M.P., Delgado, M., Pegalajar, M.C.: Multiobjetive evolutionary optimization for elman recurrent neural networks, applied to time series prediction. Fuzzy Economical Review X(1), 165–181 (2005)Google Scholar
- 10.Zitzler, E., Laumanns, M., Thiele, L.: Spea2: Improving the strength pareto evolutionary algorithm. Technical Report 103, Computer Engineering and Networks Laboratory (TIK), Swiss Federal Institute of Technology (ETH), Zurich, Switzerland (May 2001)Google Scholar