Size Of Neural Network Research Articles

ViterbiNet-Based Signal Detection for OTFS System

Recently, orthogonal time frequency space (OTFS) was presented to combat high Doppler shifts in wireless communication systems. Most studies on OTFS signal detection require channel state information (CSI). However, it is quite difficult to describe the channel model mathematically for some communication systems. This letter proposes a low-complexity ViterbiNet-based OTFS signal detection algorithm. A neural network (NN) is used in the ViterbiNet to replace the log-likelihood calculation that requires CSI in the Viterbi algorithm. Therefore, the proposed ViterbiNet-based scheme can perform signal detection without CSI in OTFS systems. Meanwhile, since it is a model-driven network, the proposed ViterbiNet-based scheme requires only a small size NN and a small amount of training data to achieve great performance. Moreover, the softplus function is utilized as the activation function, which smoothen the training of the ViterbiNet. Through experiments, simulation results prove the performance of the presented ViterbiNet-based OTFS signal detection algorithm.

Takens-inspired neuromorphic processor: A downsizing tool for random recurrent neural networks via feature extraction

We describe a new technique which minimizes the amount of neurons in the hidden layer of a random recurrent neural network (rRNN) for time series prediction. Merging Takens-based attractor reconstruction methods with machine learning, we identify a mechanism for feature extraction that can be leveraged to lower the network size. We obtain criteria specific to the particular prediction task and derive the scaling law of the prediction error. The consequences of our theory are demonstrated by designing a Takens-inspired hybrid processor, which extends a rRNN with a priori designed delay external memory. Our hybrid architecture is therefore designed including both, real and virtual nodes. Via this symbiosis, we show performance of the hybrid processor by stabilizing an arrhythmic neural model. Thanks to our obtained design rules, we can reduce the stabilizing neural network's size by a factor of 15 with respect to a standard system.

Neural Network Learning With Global Heuristic Search

A novel hybrid global optimization (GO) algorithm applied for feedforward neural networks (NNs) supervised learning is investigated. The network weights are determined by minimizing the traditional mean square error function. The optimization technique, called LP(tau)NM, combines a novel global heuristic search based on LPtau low-discrepancy sequences of points, and a simplex local search. The proposed method is initially tested on multimodal mathematical functions and subsequently applied for training moderate size NNs for solving popular benchmark problems. Finally, the results are analyzed, discussed, and compared with such as from backpropagation (BP) (Levenberg-Marquardt) and differential evolution methods.

Comparison of recent methods for inference of variable influence in neural networks

Neural networks (NNs) belong to ‘black box’ models and therefore ‘suffer’ from interpretation difficulties. Four recent methods inferring variable influence in NNs are compared in this paper. The methods assist the interpretation task during different phases of the modeling procedure. They belong to information theory (ITSS), the Bayesian framework (ARD), the analysis of the network's weights (GIM), and the sequential omission of the variables (SZW). The comparison is based upon artificial and real data sets of differing size, complexity and noise level. The influence of the neural network's size has also been considered. The results provide useful information about the agreement between the methods under different conditions. Generally, SZW and GIM differ from ARD regarding the variable influence, although applied to NNs with similar modeling accuracy, even when larger data sets sizes are used. ITSS produces similar results to SZW and GIM, although suffering more from the ‘curse of dimensionality’.