Recurrent neural networks (RNN) are a widely used tool for the prediction of time series. In this paper we use the dynamic behaviour of the RNN to categorize input sequences into different specified classes. These two tasks do not seem to have much in common. However, the prediction task strongly supports the development of a suitable internal structure, representing the main features of the in...

Frequently, sequences of state transitions are triggered by specific signals. Learning these triggered sequences with recurrent neural networks implies storing them as different attractors of the recurrent hidden layer dynamics. A challenging test and also useful for application is conditional prediction of sequences giving just the trigger signal as an input and letting the recurrent neural ne...

Limits on synaptic efficiency are characteristic of biological neural networks. In this paper, weight limitation constraints are applied to the spike time error-backpropagation (SpikeProp) algorithm for temporally encoded networks of spiking neurons. A novel solution to the problem raised by non-firing neurons is presented which makes the learning algorithm converge reliably and efficiently. In...

Central ossifying fibroma is a benign, slow-growing tumor of mesenchymal origin with a predilection for the mandibular premolar and molar areas. The immunophenotype of T cells involved in the antitumor response against this benign tumor is unknown. In this case report, we described a case of a 48-year-old woman presenting with a very large recurrent ossifying fibroma in the mandible, which was ...

This paper presents H∞ control problem for input-delayed systems. A neutral system approach is considered to the design of PD controller for input delay systems in presence of uncertain time-invariant delay. Using this approach, the resulting closed-loop system turns into a specific time-delay system of neutral type. The significant specification of this neutral system is that its delayed coeff...

The scaling of transient times to zero-lag synchronization in networks composed of excitable units is shown to be governed by three features of the graph representing the network: the longest path between pairs of neurons (diameter), the largest loop (circumference) and the loop with the maximal average out degree. The upper bound of transient times can vary between O(1) and O(N), where N is th...

This paper deals with the robust output tracking problem of uncertain large-scale systems with input delay and the time-delay interconnections. Due to the informat ion transmission between subsystems, time delays are often encountered in large-scale systems and lead to the source of system instability. First, the original nonlinear time-delay systems can be represented by the Takagi-Sugeno fuzz...

Analysis of hydrologic, geomorphic, and suburbanization data from a small instrumented drainage basin near Iowa City, Iowa, indicates that channel networks are radically altered when suburban development overtakes a drainage basin. Changes in channel networks are such that the network becomes much more efficient in collecting water quickly, so that lag time and kurtosis of storm hydrographs are...

A neural network based technique is introduced which hides the control latency of reconfigurable interconnection networks (INs) in shared memory multiprocessors. Such INs require complex control mechanisms to reconfigure the IN on demand, in order to satisfy processor-memory accesses. Hiding the control latency seen by each access improves multiprocessor performance significantly. The new techn...