18,993 Hits in 6.6 sec

A Stochastic Dynamic Local Search Method for Learning Multiple-Valued Logic Networks

2007 IEICE Transactions on Fundamentals of Electronics Communications and Computer Sciences  
In this paper, we propose a stochastic dynamic local search (SDLS) method for Multiple-Valued Logic (MVL) learning by introducing stochastic dynamics into the traditional local search method.  ...  The proposed learning network maintains some trends of quick descent to either global minimum or a local minimum, and at the same time has some chance of escaping from local minima by permitting temporary  ...  In future we plan to investigate the characteristics of a hybrid system of combining Local Search algorithm and Chaos and its application to learning Multiple-Valued Logic Networks.  ... 
doi:10.1093/ietfec/e90-a.5.1085 fatcat:mvpzslro4vhobcg2d2tv3uxlme

Learning hardware using multiple-valued logic - Part 2: Cube calculus and architecture

M. Perkowski, D. Foote, Qihong Chen, A. Al-Rabadi, L. Jozwiak
2002 IEEE Micro  
This article proposes using symbolic learning methods based on multiple-valued (MV) logic and implemented in reconfigurable hardware.  ...  For example, • for binary logic, X 1 = X, and X 0 = X ′ are two literals; and • for four-valued logic V i = {0, 1, 2, 3}: A cube on X 1 , X 2 , ... , X n is an ordered set of literals on X 1 , X 2 , ..  ...  This article proposes using symbolic learning methods based on multiple-valued (MV) logic and implemented in reconfigurable hardware.  ... 
doi:10.1109/mm.2002.1013304 fatcat:zpohw3uu3vfflaa6kzmgc6bt2i

Processing Markov Logic Networks with GPUs: Accelerating Network Grounding [chapter]

Carlos Alberto Martínez-Angeles, Inês Dutra, Vítor Santos Costa, Jorge Buenabad-Chávez
2016 Lecture Notes in Computer Science  
Markov Logic is an expressive and widely used knowledge representation formalism that combines logic and probabilities, providing a powerful framework for inference and learning tasks.  ...  Most Markov Logic implementations perform inference by transforming the logic representation into a set of weighted propositional formulae that encode a Markov network, the ground Markov network.  ...  -01-0145-FEDER-006961>>, and by National Funds through the FCT -Fundação para a Ciência e a Tecnologia (Portuguese Foundation for Science and Technology) as part of project UID/EEA/50014/2013.  ... 
doi:10.1007/978-3-319-40566-7_9 fatcat:37uwnijeavdgjifn7gs7ccf5gq

Logic Learning in Adaline Neural Network

Nadia Athirah Norani, Mohd Shareduwan Mohd Kasihmuddin, Mohd. Asyraf Mansor, Noor Saifurina Nana Khurizan
2021 Pertanika journal of science & technology  
This research incorporates logic programming that consists of various prominent logical representation. These logical rules will be a symbolic rule that defines the learning mechanism of ADNN.  ...  In this paper, Adaline Neural Network (ADNN) has been explored to simulate the actual signal processing between input and output.  ...  ADNN is a single layer network with multiple nodes where each node receives multiple inputs and produces one output.  ... 
doi:10.47836/pjst.29.1.16 fatcat:yoldhbj7xze5xc6aimydl6jklu

Ternary Logic Network Justification Using Transfer Matrices

M. A. Thornton, J. L. Dworak
2013 2013 IEEE 43rd International Symposium on Multiple-Valued Logic  
A linear algebraic method is developed that allows for logic network justification problems to be solved.  ...  The logic network is represented by a matrix that is defined as the "justification" matrix.  ...  Logic network justification is useful in multiple design and analysis applications, including synthesis, verification, and test.  ... 
doi:10.1109/ismvl.2013.57 dblp:conf/ismvl/ThorntonD13 fatcat:62qliaydsnaqncrxqrc6ywbukm

Modelling techniques for biomolecular networks [article]

Gerhard Mayer
2020 arXiv   pre-print
We also give a short overview about the mathematical frameworks for modelling of logical networks and list available software packages for logical modelling.  ...  In the end we give a short review about the difference between quantitative and qualitative models and describe the mathematics that specifically deals with qualitative modelling.  ...  Uses enhanced fast methods from computer algebra and computational algebraic geometry (rooting in Buchberger algorithm) to calculate the Gröbner bases of ideals in such rings [18] and an ideal is a set  ... 
arXiv:2003.00327v1 fatcat:ldcslhpgdrhfpavwypbx2c6qxu

Learning Algorithms via Neural Logic Networks [article]

Ali Payani, Faramarz Fekri
2019 arXiv   pre-print
We propose a novel learning paradigm for Deep Neural Networks (DNN) by using Boolean logic algebra.  ...  We demonstrate that, in contrast to the implicit learning in MLP approach, the proposed neural logic networks can learn the logical functions explicitly that can be verified and interpreted by human.  ...  allows us to manipulate the logical expressions via Algebra.  ... 
arXiv:1904.01554v1 fatcat:bkzn2i3uybelvpd4nvxkrwvyoi

Vector space weightless neural networks

Wilson Rosa de Oliveira, Adenilton J. da Silva, Teresa Bernarda Ludermir
2014 The European Symposium on Artificial Neural Networks  
By embedding the boolean space Z2 as an orthonormal basis in a vector space we can treat the RAM based neuron as a matrix (operator) acting on the vector space.  ...  weighted and quantum weightless neural models as particular cases.  ...  It is about underlying ideas and concepts. No applications and no learning algorithm are envisaged at the present work. These practical issues will be dealt with in a follow up to this paper.  ... 
dblp:conf/esann/OliveiraSL14 fatcat:eh2l6b76wzbtrlvkm3i7vu4sre

STRIP - a strip-based neural-network growth algorithm for learning multiple-valued functions

A. Ngom, I. Stojmenovic, V. Milutinovic
2001 IEEE Transactions on Neural Networks  
Preliminary experimental results are presented and discussed. Index Terms-Constructive algorithm, genetic algorithm, multiple-threshold perceptron, multiple-valued logic, neural network, partitioning.  ...  We consider the problem of synthesizing multiple-valued logic functions by neural networks. A genetic algorithm (GA) which finds the longest strip in is described.  ...  ACKNOWLEDGMENT The authors would like to thank the referees for their important and interesting suggestions.  ... 
doi:10.1109/72.914519 pmid:18244379 fatcat:ohwx3vafybeaphciww5ewsm7my

Mathematical Modeling for Network Selection in Heterogeneous Wireless Networks — A Tutorial

Lusheng Wang, Geng-Sheng G.S. Kuo
2013 IEEE Communications Surveys and Tutorials  
Index Terms-Network selection, heterogeneous wireless networks (HWNs), utility theory, multiple attribute decision making (MADM), fuzzy logic, game theory, combinatorial optimization, Markov chain.  ...  With a carefully designed unified scenario, we compare the schemes of various mathematical theories and discuss the ways to benefit from combining multiple of them together.  ...  with MADM algorithms, while some use the fuzzy logic with recursion (neural network, kernel learning, etc.).  ... 
doi:10.1109/surv.2012.010912.00044 fatcat:tlpdpnlzjbbefon4mbrhfznvwu

Comparing Logic Programming in Radial Basis Function Neural Network (RBFNN) and Hopfield Neural Network

Mamman Mamuda, Saratha Sathasivam
2014 International journal of computational and electronics aspects in engineering  
Neural network is a black box that clearly learns the internal relations of unknown systems. Neural-symbolic systems are based on both logic programming and artificial neural networks.  ...  This study gives an overview of how logic programming is been carried out on both networks as well as the comparison of doing logic programming on both radial basis neural network and Hopfield neural network  ...  Every learning algorithm of perceptron's for Hopfield network can be turned into a learning method.  ... 
doi:10.26706/ijceae.1.1.20141204 fatcat:lqvbyk45u5g47ki4qqkgpwdch4

A Unified Programmable Edge Matrix Processor for Deep Neural Networks and Matrix Algebra

Biji George, Om ji Omer, Ziaul Choudhury, Anoop V, Sreenivas Subramoney
2022 ACM Transactions on Embedded Computing Systems  
Matrix Algebra and Deep Neural Networks represent foundational classes of computational algorithms across multiple emerging applications like Augmented Reality(AR) or Virtual Reality(VR), autonomous navigation  ...  We submit MxCore as the generalized approach to facilitate the flexible acceleration of multiple Matrix Algebra and Deep-learning applications across a range of sparsity levels.  ...  INTRODUCTION Emerging computer applications of Artiicial Intelligence [63, 70] based on Deep learning [49] and other Machine learning [45] based approaches use computationally intensive tensor algebra  ... 
doi:10.1145/3524453 fatcat:miqhwzep3fey5admehib4md5ly

On Properties of Networks of Neuron-Like Elements

Pierre Baldi, Santosh S. Venkatesh
1987 Neural Information Processing Systems  
The dual problem of determining the computational capacity of a class of multi-layered networks with dynamics regulated by an algebraic Hamiltonian is considered.  ...  Some conclusions are also drawn about learning complexity, and some open problems raised.  ...  Theorem 1 The maximal (algorithm independent) storage capacity of a homogeneous algebraic threshold network of degree d is less than or equal to 2 ( ~ ).  ... 
dblp:conf/nips/BaldiV87 fatcat:ucj2sk3625dalnjkg7ghpuflhq

Understanding Neural Networks for Machine Learning using Microsoft Neural Network Algorithm

Nagesh Ramprasad
2016 International Journal of Computer Applications  
The Microsoft Neural System Algorithm is a simple implementation of the adaptable and popular neural networks that are used in the machine learning.  ...  A Neural Network refers to a simple computing system that consists of some highly fixed and interconnected processing elements. All the neural networks appear as a set of layers.  ...  Neural networks have a learning rate that can be approximated as n=0.1 and a regulation parameter of about x=0.5.  ... 
doi:10.5120/ijca2016911481 fatcat:3okv4qgwlbg5njih5fkd7e5q2i

Array Languages Make Neural Networks Fast [article]

Artjoms Šinkarovs, Hans-Nikolai Vießmann, Sven-Bodo Scholz
2019 arXiv   pre-print
Modern machine learning frameworks are complex: they are typically organised in multiple layers each of which is written in a different language and they depend on a number of external libraries, but at  ...  We do this by implementing a state of the art Convolutional Neural Network (CNN) and compare it against implementations in TensorFlow and PyTorch --- two state of the art industrial-strength frameworks  ...  As a result modern networks require advanced and powerful hardware -modern machine learning applications are envisioned to run on massively parallel high-throughput systems that may be equipped with GPUs  ... 
arXiv:1912.05234v1 fatcat:mb7pjcrg2jbxxiduto4jrv2nca
« Previous Showing results 1 — 15 out of 18,993 results