How Neural Nets Work Alan Lapedes Robert Farber Theoretical Division Los Alamos National Laboratory Los Alamos, NM 87545 There is presently great interest in the abilities of neural networks to mimic "qualitative reasoning" by manipulating neural incodings of symbols. Less work has been performed on using neural networks to process floating point numbers and it is sometimes stated that neural n...

In this paper, we introduce transformations of deep rectifier networks, enabling the conversion of deep rectifier networks into shallow rectifier networks. We subsequently prove that any rectifier net of any depth can be represented by a maximum of a number of functions that can be realized by a shallow network with a single hidden layer. The transformations of both deep rectifier nets and deep...

Cloud and cloud shadow segmentation are fundamental processes in optical remote sensing image analysis. Current methods for cloud/shadow identification geospatial imagery not as accurate they should, especially the presence of snow haze. This paper presents a deep learning-based framework detection Landsat 8 images. Our method benefits from convolutional neural network, Cloud-Net+ (a modificati...

We consider propagation of coherent light through a nonlinear periodic optical structure consisting of two alternating layers with different linear and nonlinear refractive indices. A coupled-mode system is derived from the Maxwell equations and analyzed for the stationary-transmission regimes and linear time-dependent dynamics. We find the domain for existence of true all-optical limiting when...

It has been known for quite a while that the Vapnik-Chervonenkis dimension (VC-dimension) of a feedforward neural net with linear threshold gates is at most O(w . log w), where w is the total number of weights in the neural net. We show in this paper that this bound is in fact asymptotically optimal. More precisely, we exhibit for any depth d 2 3 a large class of feedforward neural nets of dept...

Spiking Neural P (SN P) system characterizes the movement of spikes among the neurons. Spikes have a similarity with tokens in Petri net where tokens (like spikes) are moved through net according to specific rules. This paper proposes the concept of spiking Petri nets, which are isomorphic to spiking neural P systems. It also gives algorithms to construct spiking Petri net for SN P system and v...

Applications of complex valued neural networks to optical signal processing using phase sensitive detection schemes are proposed and discussed In optical information processing systems the advantages of complex valued neural networks are realized directly utilizing physical phenomena of lightwave and phase sensitive optical detection In the proposed system the summation and nonlinear transform ...

In modern day pattern recognition, neural nets are used extensively. General use of a feedforward neural net consists of a training phase followed by a classi cation phase. Classication of an unknown test vector is very fast and only consists of the propagation of the test vector through the neural net. Training involves an optimization procedure and is very time consuming since a feasible loca...

There exist a number of negative results ([J), [BR), [KV]) about learning on neural nets in Valiant's model [V) for probably approximately correct learning ("PAC-learning"). These negative results are based on an asymptotic analysis where one lets the number of nodes in the neural net go to infinit.y. Hence this analysis is less adequate for the investigation of learning on a small fixed neural...

In this paper, we propose a model using generative adversarial net (GAN) to generate realistic text. Instead of using standard GAN, we combine variational autoencoder (VAE) with generative adversarial net. The use of high-level latent random variables is helpful to learn the data distribution and solve the problem that generative adversarial net always emits the similar data. We propose the VGA...