536,489 Hits in 4.0 sec

Constructive neural network learning [article]

Shaobo Lin, Jinshan Zeng, Xiaoqin Zhang
2016 arXiv   pre-print
FNNs learning system called the constructive feed-forward neural network (CFN).  ...  Motivated by the idea of "constructive neural networks" in approximation theory, we focus on "constructing" rather than "training" feed-forward neural networks (FNNs) for learning, and propose a novel  ...  We concluded in this paper by presenting some extensions of the constructive neural networks learning. In the present paper, the neural network was constructed by using the method in [24] .  ... 
arXiv:1605.00079v1 fatcat:mw6pgxishvah5h2z4ypzfk3wce

Continuously Constructive Deep Neural Networks [article]

Ozan İrsoy, Ethem Alpaydın
2018 arXiv   pre-print
Traditionally, deep learning algorithms update the network weights whereas the network architecture is chosen manually, using a process of trial and error.  ...  We propose two methods: In tunnel networks, this selection is done at the level of a hidden unit, and in budding perceptrons, this is done at the level of a network layer; updating this control parameter  ...  For future work, it will be interesting to see different application areas of our constructive neural networks.  ... 
arXiv:1804.02491v1 fatcat:uaaop2mipfbjtijvb3cawh3xzu

Constructive Neural Network: A Framework

2019 International Journal of Engineering and Advanced Technology  
In this paper, two techniques for construction of feedforward neural network are being reviewed: pruning neural network algorithms and constructive neural network algorithms.  ...  A number of major issues are discussed that can be considered while constructing a constructive neural network i.e. how to select network architecture, network growing strategy, weight freezing, optimization  ...  The adaptive architecture algorithms are of two types: pruning neural network and constructive neural network.  ... 
doi:10.35940/ijeat.b3304.129219 fatcat:c3g3k5ofo5c75dzdx2ovfusj3q

Constructing Neural Circuits and Networks [chapter]

Michael Vanier, David Beeman
1998 The Book of GENESIS  
This will not be a "neural network" in the usual sense of a network of highly abstract units with no direct connection to biological neurons (such as a backpropagation network).  ...  In this chapter, we demonstrate how to use GENESIS to set up a simple network of biologically realistic neurons.  ...  Constructing Neural Circuits and Networks & @ ¢ V % F ¡ E U ¦ ¥ ¦ ' £ ( & @ G © ¥ V U ¦ U ¦ E ¥ % T U § A £ ¡ ) $ ¥ W A £ ¡ ( & @ £ W F % © £ % ¤ § $ E ¥ ¦ X A ¥ W H ( & @ ¥ $ U F V § © £ ¡ © ¥ E % A ¤  ... 
doi:10.1007/978-1-4612-1634-6_18 fatcat:oxzebweqqjgxxfqwzhlsvufc2y

How to Construct Deep Recurrent Neural Networks [article]

Razvan Pascanu, Caglar Gulcehre, Kyunghyun Cho, Yoshua Bengio
2014 arXiv   pre-print
In this paper, we explore different ways to extend a recurrent neural network (RNN) to a deep RNN.  ...  We start by arguing that the concept of depth in an RNN is not as clear as it is in feedforward neural networks.  ...  Deep Recurrent Neural Networks Why Deep Recurrent Neural Networks?  ... 
arXiv:1312.6026v5 fatcat:dkmi2dyijncjlhnqlpay3klnrm

Applying Artificial Neural Networks In Construction

Anna Doroshenko, P. Zhou, Y. He, R. Weerasinghe
2020 E3S Web of Conferences  
Currently, artificial neural networks (ANN) are used to solve the following complex problems: pattern recognition, speech recognition, complex forecasts and others.  ...  This paper presents an overview of applications of ANN in construction industry, including energy efficiency and energy consumption, structural analysis, construction materials, smart city and BIM technologies  ...  neural networks -networks of nerve cells of a living organism.  ... 
doi:10.1051/e3sconf/202014301029 fatcat:3d4ri3ikjnbr5iwuwtwxhuecua

Evolutionary Construction of Convolutional Neural Networks [article]

Marijn van Knippenberg, Vlado Menkovski, Sergio Consoli
2019 arXiv   pre-print
It combines deep neural networks and evolutionary algorithms to improve and/or automate the construction of neural networks.  ...  Recent Neuro-Evolution approaches have shown promising results, rivaling hand-crafted neural networks in terms of accuracy.  ...  Constructing a neural network is still often seen as a somewhat "magic" skill by many: a combination of knowledge, past experience, and intuition.  ... 
arXiv:1903.01895v1 fatcat:hyhajkfoerf6he67jmws3q3fly


Nahm-Woo Hahm, Bum-Il Hong
2012 Honam Mathematical Journal  
In this paper, we discuss a constructive approximation by Gaussian neural networks.  ...  We show that it is possible to construct Gaussian neural networks with integer weights that approximate arbitrarily well for functions in Cc(R s ).  ...  In addition, we are not able to use the constructive proofs in [2] and [6] if the activation function in a neural network is not a sigmoidal function.  ... 
doi:10.5831/hmj.2012.34.3.341 fatcat:d6wpobotafahhnjevfbsn7e344

Constructive neural networks: some practical considerations

Tin-Yau Kwok, Dit-Yan Yeung
1994 Proceedings of 1994 IEEE International Conference on Neural Networks (ICNN'94)  
Abstract| Based on a Hilbert space point of view, we proposed in our previous work a novel objective function for training new hidden units in a constructive feedforward neural network.  ...  Moreover, we p r o ved that if the hidden unit functions satisfy the universal approximation property, the network so constructed incrementally, using the proposed objective function and with input weight  ...  In 12], we formulated the problem of learning in constructive neural networks as constructive a pproximation in a Hilbert space.  ... 
doi:10.1109/icnn.1994.374162 fatcat:p2mygofmvnhzhpay7s4rr5emxu

Constructive approximate interpolation by neural networks

B. Llanas, F.J. Sainz
2006 Journal of Computational and Applied Mathematics  
We present a type of single-hidden layer feedforward neural networks with sigmoidal nondecreasing activation function. We call them ai-nets.  ...  All these capabilities are based on a closed expression of the networks.  ...  Acknowledgements The authors wish to thank the reviewers for their constructive suggestions and comments.  ... 
doi:10.1016/ fatcat:ahip7kpxmfh45n7zjopl7etnty

Constructive training of probabilistic neural networks

Michael R. Berthold, Jay Diamond
1998 Neurocomputing  
This paper presents an easy to use, constructive training algorithm for Probabilistic Neural Networks, a special type of Radial Basis Function Networks.  ...  It is demonstrated that the proposed algorithm generates Probabilistic Neural Networks that achieve a comparable classi cation performance on these datasets.  ...  Probabilistic Neural Networks The Probabilistic Neural Network was introduced in 1990 by S p e c ht 17 and puts the statistical kernel estimator 8 into the framework of Radial Basis Function Networks.  ... 
doi:10.1016/s0925-2312(97)00063-5 fatcat:7prbtbflbvg7ffhg64jjpnlaxy

Neural Network-based Automatic Factor Construction [article]

Jie Fang, Jianwu Lin, Shutao Xia, Yong Jiang, Zhikang Xia, Xiang Liu
2020 arXiv   pre-print
This paper proposes Neural Network-based Automatic Factor Construction (NNAFC), a tailored neural network framework that can automatically construct diversified financial factors based on financial domain  ...  knowledge and a variety of neural network structures.  ...  Contributions In this paper, a novel network structure called Neural Network-based Automatic Factor Construction (NNAFC) is proposed, which can use deep neural networks to automatically construct financial  ... 
arXiv:2008.06225v3 fatcat:atwe4tumzzgolls76qgw65nvii

Constructing hysteretic memory in neural networks

Jyh-Da Wei, Chuen-Tsai Sun
2000 IEEE Transactions on Systems Man and Cybernetics Part B (Cybernetics)  
Index Terms-Hysteresis, hysteretic memory, rate independence, recurrent network, reinforcement learning, time delay neural network.  ...  In addition to other memoryrelated studies such as time delay neural networks, recurrent networks, and reinforcement learning, rate-independent memory deserves further attention owing to its potential  ...  Network computation can be more fully implemented to achieve a better performance from those systems that are hysteresis embedded if neural networks could be constructed as hysteresis simulators.  ... 
doi:10.1109/3477.865179 pmid:18252392 fatcat:b2x6qcemjrf25bcdmnnlu5x76m


Smita K Magdum, Amol C Adamuthe
2017 ICTACT Journal on Soft Computing  
The objective of this paper is to develop neural networks and multilayer perceptron based model for construction cost prediction.  ...  Four artificial neural network models and twelve multilayer perceptron models are compared. MLP and NN give better results than statistical regression method.  ...  CONSTRUCTION COST PREDICTION MODEL This section describes neural network based approach for construction cost prediction.  ... 
doi:10.21917/ijsc.2017.0216 fatcat:n5ahqe6p6nbmbdilszaqzmm5oq

Construction Project Performance Model Using Artificial Neural Network

2017 International Journal of Recent Trends in Engineering and Research  
Objective of this research is to develop Artificial neural network (ANN) models to predict cost performance, schedule performance, quality performance and satisfaction level.  ...  Success of construction projects depends mainly on success of performance of a project.  ...  LITERATURE REVIEW Construction projects are dynamic and time-consuming undertakings. Ling and Liu [1] predicted the performance of projects using neural networks.  ... 
doi:10.23883/ijrter.2017.3199.deyet fatcat:7wnksx4lirabddsxg3gch3edv4
« Previous Showing results 1 — 15 out of 536,489 results