Locally recurrent globally feedforward networks: a critical reviewof architectures

In this paper, we will consider a number of local-recurrent-global-feedforward (LRGF) networks that have been introduced by a number of research groups in the past few years. We first analyze the various architectures, with a view to highlighting their differences. Then we introduce a general LRGF network structure that includes most of the network architectures that have been proposed to date. Finally we will indicate some open issues concerning these types of networks.     

Multilayer perceptron trained using radial basis functions

An algebraic equation for the training of a multilayer perceptron using radial basis functions is derived. Examples using this technique for the training of a network for the exclusive-OR and related problems are presented. Suggestions on the choice of the number of centres are given.<>     

Adaptive prediction of an arma lattice

In the letter, based on a modification of a well-known approximate likelihood estimation technique, a one-step-ahead prediction method using an ARMA lattice structure is obtained.     

An evaluation of the neocognitron

We describe a sequence of experiments investigating the strengths and limitations of Fukushima's neocognitron as a handwritten digit classifier. Using the results of these experiments as a foundation, we propose and evaluate improvements to Fukushima's original network in an effort to obtain higher recognition performance. The neocognitron performance is shown to be strongly dependent on the choice of selectivity parameters and we present two methods to adjust these variables. Performance of the network under the more effective of the two new selectivity adjustment techniques suggests that the network fails to exploit the features that distinguish different classes of input data. To avoid this shortcoming, the network's final layer cells were replaced by a nonlinear classifier (a multilayer perceptron) to create a hybrid architecture. Tests of Fukushima's original system and the novel systems proposed in this paper suggest that it may be difficult for the neocognitron to achieve the performance of existing digit classifiers due to its reliance upon the supervisor's choice of selectivity parameters and training data.     

Experimental investigations of blast-induced consolidation of loess soils

A method of consolidating loess soils by the detonation of horizontal explosive charges situated in two tiers at a distance of 3–4 m with respect to depth is proposed. Test results for a different alternation of the detonation of horizontal charges in the upper and lower tiers are cited. The saving generated by the method adopted is indicated. Translated from Osnovaniya, Fundamenty i Mekhanika Gruntov, No. 6, pp. 9–13, November–December, 1994.