Training feedforward networks with the Marquardt algorithm

By Martin T. Hagan, M. Menhaj. 1994

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

1. Gong Cheng
2. Junwei Han
3. Xiwen Yao
4. Shiji Song
5. Peicheng Zhou
6. Gao Huang
7. Luis Gómez-Chova
8. Lei Guo

Citations in the corpus

• An Activity Recognition System For Mobile Phones (2009, methodology): “This method is a very accurate learning algorithm [ 8 ], but due to its O(n 2 ) memory and speed requirements it can only be”
• A comprehensive review for industrial applicability of artificial neural networks (2003, methodology): “Although the back-propagation algorithm has been a significant milestone, many attempts have been made to speed up the convergence and significant improvement are observed by using various second order approaches, namely, Newton’s method, conjugate gradient’s, or the LM optimization technique [5], [”
• Freeway Travel Time Prediction with State-Space Neural Networks: Modeling State-Space Dynamics with Recurrent Neural Networks (2002, background): “For a comprehensive treatment of these techniques, see elsewhere ( 23, 24 ).”
• A hybrid model for high-frequency stock market forecasting (2015, methodology): “…MLP layers, since this configuration achieves improved prediction performance (de A. Araújo, 2007), as well as the Levenberg Marquardt algorithm (Hagan & Menhaj, 1994) for the MLP learning process with the following stop criteria (Prechelt, 1994): (i) epoch ¼ 10;000, (ii) Pt 6 10 6, and (iii)…”
• Performance evaluation of two different neural network and particle swarm optimization methods for prediction of discharge capacity of modified triangular side weirs (2014, methodology): “It is one of the most useful algorithms in MLP neural network [38].”
• Differential Evolution Training Algorithm for Feed-Forward Neural Networks (2003, methodology): “Widely applied methods are,for example,modified back-propagation [11],backpropagation using the conjugate-gradient approach [2],scaled conjugate-gradient and its stochastic counterpart [14],the Marquadt algorithm [ 7 ],and a concept learning based back-propagation [8].”
• Forecasting SO 2 Pollution Incidents by means of Elman Artificial Neural Networks and ARIMA Models (2013, methodology): “The Levenberg-Marquardt algorithm, whose application is currently very common for RNN, was chosen as it does not suffer the slow convergence problems that were reported in the methods [18] from which it is derived.”