Preventing local minima by decorrelation

Abstract

In this paper, a new method of identifying and eliminating local minima which commonly appear in recurrent neural networks is presented. The energy surface of the neural network is re-sculptured by decorrelating the spurious states during learning process so as to remove the local minima. The spurious states are identified by applying a stationary condition to the set of admissible states. The stability verification is done efficiently by a specially designed parallel network. The decorrelation is employed to only those predetermined spurious states. As the result of this "unlearning", the correct retrieval rate as well as the storage capacity of the dynamic associative memory is improved.

Related Papers

• → A Comparison between Horizontal and Vertical Interchannel Decorrelation(2017)3 cited
• → Decorrelation of intravascular ultrasound signals: a computer simulation study(2002)1 cited
• → Effect of signal decorrelation in channels of phased antenna arrays on the efficiency of adaptive spatial filtering(2017)1 cited
• The Perceptual Effects of Horizontal and Vertical Interchannel Decorrelation, using the Lauridsen Decorrelator(2014)
• → Removing Spurious Features can Hurt Accuracy and Affect Groups Disproportionately(2020)3 cited