Student or educator? Sign up with your school email and get 6 months of Pro free.
CoChat

The DOI matches its Crossref record; 52 of the 52 references it cites were found in a registry or an index.

Scalable training of artificial neural networks with adaptive sparse connectivity inspired by network science

Open access
Mocanu, Decebal Constantin, Mocanu, Elena, Stone, Peter, Nguyen, Phuong H., Gibescu, Madeleine, Liotta, Antonio · 2018 · Nature Communications

Abstract

Through the success of deep learning in various domains, artificial neural networks are currently among the most used artificial intelligence methods. Taking inspiration from the network properties of biological neural networks (e.g. sparsity, scale-freeness), we argue that (contrary to general practice) artificial neural networks, too, should not have fully-connected layers. Here we propose sparse evolutionary training of artificial neural networks, an algorithm which evolves an initial sparse topology (Erdős–Rényi random graph) of two consecutive layers of neurons into a scale-free topology, during learning. Our method replaces artificial neural networks fully-connected layers with sparse ones before training, reducing quadratically the number of parameters, with no decrease in accuracy. We demonstrate our claims on restricted Boltzmann machines, multi-layer perceptrons, and convolutional neural networks for unsupervised and supervised learning on 15 datasets. Our approach has the potential to enable artificial neural networks to scale up beyond what is currently possible.

Overview

The claims and statistics of this paper haven’t been checked yet.

Free: it reads the open-access text and checks the first 5 claims that cite a source we can check. Create a free account to check every claim.

A few free checks a day · papers already verified open instantly · results are public

Integrity signals

Registry

Crossref · 10.1038/s41467-018-04316-3

Cites

52 of 52 references verified0 citation errors · 0 not checkable

Statistics

Figures

Fig. 1 An illustration of the SET procedure. For each sparse connected layer, SCk (a), of an ANN at the end of a training epoch a fraction of the weights, the ones closest to zero, are removed (b). Then, new weighs are added randomly in the same amount as the ones previously removed (c). Further on, a new training epoch is performed (d), and the procedure to remove and add weights is repeated. The process continues for a finite number of training epochs, as usual in the ANNs training
Fig. 1 An illustration of the SET procedure. For each sparse connected layer, SCk (a), of an ANN at the end of a training epoch a fraction of the weights, the ones closest to zero, are removed (b). Then, new weighs are added randomly in the same amount as the ones previously removed (c). Further on, a new training epoch is performed (d), and the procedure to remove and add weights is repeated. The process continues for a finite number of training epochs, as usual in the ANNs trainingCC BY · PubMed Central
Algorithm 1: SET pseudocode
Algorithm 1: SET pseudocodeCC BY · PubMed Central
Fig. 2 Experiments with RBM variants on the DNA dataset. For each model studied we have considered three cases for the number of contrastive divergence steps, nCD = 1 (a–c), nCD = 3 (d–f), and nCD = 10 (g–i). Also, we considered three cases for the number of hidden neurons, nh = 100 (a,d,g), nh = 250 (b,e,h), and nh = 500 (c,f,i). In each panel, the x axes show the training epochs; the left y axes show the average log-probabilities computed on the test data with AIS33; and the right y axes (the stacked bar on the right side of the panels) reflect the fraction given by the nW of each model over the sum of the nW of all three models. Overall, SET-RBM outperforms the other two models in most of the cases. Also, it is interesting to see that SET-RBM and RBMFixProb are much more stable and do not present the over-fitting problems of RBM
Fig. 2 Experiments with RBM variants on the DNA dataset. For each model studied we have considered three cases for the number of contrastive divergence steps, nCD = 1 (a–c), nCD = 3 (d–f), and nCD = 10 (g–i). Also, we considered three cases for the number of hidden neurons, nh = 100 (a,d,g), nh = 250 (b,e,h), and nh = 500 (c,f,i). In each panel, the x axes show the training epochs; the left y axes show the average log-probabilities computed on the test data with AIS33; and the right y axes (the stacked bar on the right side of the panels) reflect the fraction given by the nW of each model over the sum of the nW of all three models. Overall, SET-RBM outperforms the other two models in most of the cases. Also, it is interesting to see that SET-RBM and RBMFixProb are much more stable and do not present the over-fitting problems of RBMCC BY · PubMed Central
Fig. 3 SET-RBM evolution towards a scale-free topology on the DNA dataset. We have considered three cases for the number of contrastive divergence steps, nCD = 1 (a–c), nCD = 3 (d–f), and nCD = 10 (g–i). Also, we considered three cases for the number of hidden neurons, nh = 100 (a, d, g), nh = 250 (b, e, h), and nh = 500 (c, f, i). In each panel, the x axes show the training epochs; the left y axes (red color) show the average log-probabilities computed for SET-RBMs on the test data with AIS33; and the right y axes (cyan color) show the p-values computed between the degree distribution of the hidden neurons in SET-RBM and a power-law distribution. We may observe that for models with a high enough number of hidden neurons, the SET-RBM topology always tends to become scale-free
Fig. 3 SET-RBM evolution towards a scale-free topology on the DNA dataset. We have considered three cases for the number of contrastive divergence steps, nCD = 1 (a–c), nCD = 3 (d–f), and nCD = 10 (g–i). Also, we considered three cases for the number of hidden neurons, nh = 100 (a, d, g), nh = 250 (b, e, h), and nh = 500 (c, f, i). In each panel, the x axes show the training epochs; the left y axes (red color) show the average log-probabilities computed for SET-RBMs on the test data with AIS33; and the right y axes (cyan color) show the p-values computed between the degree distribution of the hidden neurons in SET-RBM and a power-law distribution. We may observe that for models with a high enough number of hidden neurons, the SET-RBM topology always tends to become scale-freeCC BY · PubMed Central
Fig. 4 SET-RBMs connectivity patterns for the visible neurons. a On the MNIST dataset. b On the Caltech 101 16 × 16 dataset. For each dataset, we have analyzed two SET-RBM architectures, i.e. 500 and 2500 hidden neurons. The heat-map matrices are obtained by reshaping the visible neurons vector to match the size of the original input images. In all cases, it can be observed that the connectivity starts from an initial Erdös–Rényi distribution. Then, during the training process, it evolves towards organized patterns which depend on the input images
Fig. 4 SET-RBMs connectivity patterns for the visible neurons. a On the MNIST dataset. b On the Caltech 101 16 × 16 dataset. For each dataset, we have analyzed two SET-RBM architectures, i.e. 500 and 2500 hidden neurons. The heat-map matrices are obtained by reshaping the visible neurons vector to match the size of the original input images. In all cases, it can be observed that the connectivity starts from an initial Erdös–Rényi distribution. Then, during the training process, it evolves towards organized patterns which depend on the input imagesCC BY · PubMed Central
Fig. 5 Experiments with MLP variants using three benchmark datasets. a,c,e reflect models performance in terms of classification accuracy (left y axes) over training epochs (x axes); the right y axes of a,c,e give the p-values computed between the degree distribution of the hidden neurons of the SET-MLP models and a power-law distribution, showing how the SET-MLP topology becomes scale-free over training epochs. b,d,f depict the number of weights of the three models on each dataset. The most striking situation happens for the CIFAR10 dataset (c,d) where the SET-MLP model outperforms drastically the MLP model, while having ~100 times fewer parameters
Fig. 5 Experiments with MLP variants using three benchmark datasets. a,c,e reflect models performance in terms of classification accuracy (left y axes) over training epochs (x axes); the right y axes of a,c,e give the p-values computed between the degree distribution of the hidden neurons of the SET-MLP models and a power-law distribution, showing how the SET-MLP topology becomes scale-free over training epochs. b,d,f depict the number of weights of the three models on each dataset. The most striking situation happens for the CIFAR10 dataset (c,d) where the SET-MLP model outperforms drastically the MLP model, while having ~100 times fewer parametersCC BY · PubMed Central
Fig. 6 Models accuracy using three weights regularization techniques on the Fashion-MNIST dataset. All models have been trained with stochastic gradient descent, having the same hyper-parameters, number of hidden layers (i.e. three), and number of hidden neurons per layer (i.e. 1000). a–c use ReLU activation function for the hidden neurons and Nesterov momentum; d–f use ReLU activation function without Nesterov momentum; g–i use SReLU activation function and Nesterov momentum; and j–l use SReLU activation function without Nesterov momentum. a,d,g,j present experiments with SET-MLP; b,e,h,k with MLPFixProb; and c,f,i,l with MLP
Fig. 6 Models accuracy using three weights regularization techniques on the Fashion-MNIST dataset. All models have been trained with stochastic gradient descent, having the same hyper-parameters, number of hidden layers (i.e. three), and number of hidden neurons per layer (i.e. 1000). a–c use ReLU activation function for the hidden neurons and Nesterov momentum; d–f use ReLU activation function without Nesterov momentum; g–i use SReLU activation function and Nesterov momentum; and j–l use SReLU activation function without Nesterov momentum. a,d,g,j present experiments with SET-MLP; b,e,h,k with MLPFixProb; and c,f,i,l with MLPCC BY · PubMed Central
Fig. 7 Experiments with CNN variants on the CIFAR10 dataset. a Models performance in terms of classification accuracy (left y axes) over training epochs (x axes). b The number of weights of the three models on each dataset. The convolutional layers of each model have in total 287,008 weights, while the fully connected (or the sparse) layers on top have 8,413,194, 184.842, and 184,842 weights for CNN, CNNFixProb, and SET-CNN, respectively
Fig. 7 Experiments with CNN variants on the CIFAR10 dataset. a Models performance in terms of classification accuracy (left y axes) over training epochs (x axes). b The number of weights of the three models on each dataset. The convolutional layers of each model have in total 287,008 weights, while the fully connected (or the sparse) layers on top have 8,413,194, 184.842, and 184,842 weights for CNN, CNNFixProb, and SET-CNN, respectivelyCC BY · PubMed Central

Verification

Claims the paper makes

The claims and statistics of this paper haven’t been checked yet.

Free: it reads the open-access text and checks the first 5 claims that cite a source we can check. Create a free account to check every claim.

A few free checks a day · papers already verified open instantly · results are public

Claims other papers make about this one

Each is checked against this paper’s own text: does it say what the citing paper says it does?

Self-check

Percentages, sums and reported differences re-computed from the paper’s own numbers.

The self-check needs the paper’s full text.

Citations

Cites

Show all 45 references

Research works printed but not linked: 19

  1. Mocanu, D. C. On the synergy of network science and artificial intelligence. In Proc. 25th International Joint Conference on Artificial Intelligence(ed. Kambhampati, S.) 4020–4021 (AAAI Press, New York, 2016).
  2. Dieleman, S. & Schrauwen, B. Accelerating sparse restricted boltzmann machine training using non-gaussianity measures. In Proc. Deep Learning and Unsupervised Feature Learning, Vol. 9 (eds Bengio Y., Bergstra J. & Le Q.) http://hdl.handle.net/1854/LU-3118568 (Lake Tahoe, 2012).
  3. Yosinski, J. & Lipson, H. Visually debugging restricted boltzmann machine training with a 3d example. In Representation Learning Workshop, 29th International Conference on Machine Learning (Edinburgh, 2012).
  4. Whiteson, S. & Stone, P. Evolutionary function approximation for reinforcement learning. J. Mach. Learn. Res. 7, 877–917 (2006).
  5. Weigend, A. S., Rumelhart, D. E. & Huberman, B. A. Generalization by weight-elimination with application to forecasting. In Proc. Advances in Neural Information Processing Systems, Vol. 3, 875–882 (Morgan-Kaufmann, Colorado, 1991).
  6. Smolensky, P. Information processing in dynamical systems: foundations of harmony theory. In Parallel Distributed Processing: Explorations in the Microstructure of Cognition (eds Rumelhart, D. E., McClelland, J. L. & CORPORATE PDP Research Group) 194–281 (MIT Press, Cambridge, 1986).
  7. Osogami, T. & Otsuka, M. Restricted boltzmann machines modeling human choice. Proc. Adv. Neural Inf. Process. Syst. 27, 73–81 (2014).
  8. Everitt, B. The Cambridge Dictionary of Statistics (Cambridge University Press, Cambridge, UK; New York, 2002).
  9. Al-Rfou, R., et al. Theano: a Python framework for fast computation of mathematical expressions. Preprint at https://arxiv.org/abs/1605.02688 (2016). · arXiv
  10. Urban, G. et al. Do deep convolutional nets really need to be deep and convolutional? In Proc. 5th International Conference on Learning Representations (OpenReview.net, Toulon, 2016).
  11. Nair, V. & Hinton, G. E. Rectified linear units improve restricted boltzmann machines. In Proc. 27th International Conference on Machine Learning (eds Fürnkranz, J. & Joachims, T.) 807–814 (Omnipress, Haifa, 2010).
  12. Barabási, A.-L. Network Science (Cambridge University Press, Glasgow, 2016).
  13. Bishop, C. M. Pattern Recognition and Machine Learning (Information Science and Statistics) (Springer-Verlag New York, Inc., Secaucus, 2006).
  14. Sutton, R. S. & Barto, A. G. Introduction to Reinforcement Learning. (MIT Press, Cambridge, MA, USA, 1998).
  15. Bottou, L. & Bousquet, O. The tradeoffs of large scale learning. In Proc. Advances in Neural Information Processing Systems Vol. 20 (eds Platt, J. C., Koller, D., Singer, Y. & Roweis, S. T.) 161–168 (NIPS Foundation, Vancouver, 2008).
  16. Larochelle, H. & Murray, I. The neural autoregressive distribution estimator. In Proc. 14th International Conference on Artificial Intelligence and Statistics(eds Gordon, G., Dunson, D. & Dudík, M.) 29–37 (JMLR, Fort Lauderdale, 2011).
  17. Marlin, B. M., Swersky, K., Chen, B. & de Freitas, N. Inductive principles for restricted boltzmann machine learning. In Proc. 13th International Conference on Artificial Intelligence and Statistics (eds Teh, Y. W. & Titterington, M.) 509–516 (JMLR, Sardinia, 2010).
  18. LeCun, Y., Bottou, L., Bengio, Y. & Haffner, P. Gradient-based learning applied to document recognition. Proc. IEEE 86, 2278–2324 (1998). · DOI · Paper page
  19. Krizhevsky, A. Learning Multiple Layers of Features from Tiny Images. Master’s thesis (2009).

Cited by

See all 598

Metadata

Authors
Mocanu, D. C., Mocanu, E., Stone, P., Nguyen, P. H., Gibescu, M., & Liotta, A.
Affiliations
  • Eindhoven University of Technology (NL)
  • The University of Texas at Austin (US)
  • University of Derby (GB)
Corresponding author
Mocanu, Decebal Constantin
Venue
Nature Communications
Published
2018-06-19
Volume
9
Issue
1
Pages
2383-2383
ISSN
2041-1723 (ISSN-L)
OpenAlex
W2808133870
Cited by
524
References
45
Citations by year
2026: 45 · 2025: 85 · 2024: 72 · 2023: 68 · 2022: 60 · 2021: 100 · 2020: 51 · 2019: 39 · 2018: 4
FWCI
27.30
Type
journal article
Language
en
Open access
Licence
Locations
  • Nature Communications (publishedVersion, cc-by)
  • Europe PMC (PubMed Central) (publishedVersion, other-oa)
  • PubMed Central (publishedVersion, cc-by)
  • PubMed (publishedVersion)
  • arXiv (Cornell University) (submittedVersion)
  • Open Repository and Bibliography (University of Luxembourg) (submittedVersion, other-oa)
  • TU/e Research Portal (submittedVersion, cc-by)
  • TU/e Research Portal (submittedVersion)
  • RePEc: Research Papers in Economics (submittedVersion)
  • DOAJ (DOAJ: Directory of Open Access Journals) (submittedVersion, cc-by-sa)
  • DOAJ (DOAJ: Directory of Open Access Journals) (submittedVersion)
  • Utrecht University Repository (Utrecht University) (submittedVersion, other-oa)
  • Research Output (Edinburgh Napier University) (submittedVersion, cc-by)
  • Edinburgh Napier Research Repository (Edinburgh Napier University) (submittedVersion, cc-by)
  • Open Research (University of Surrey) (submittedVersion, other-oa)
  • arXiv (Cornell University)
Topics
  • Physical Sciences › Computer Science › Artificial Intelligence › Neural Networks and Applications
  • Physical Sciences › Computer Science › Computer Vision and Pattern Recognition › Advanced Neural Network Applications
  • Physical Sciences › Physics and Astronomy › Statistical and Nonlinear Physics › Complex Network Analysis Techniques
Keywords
sparse neural networks, scale-free topology, Erdős-Rényi random graphs, parameter reduction, restricted Boltzmann machine, multilayer perceptron, convolutional neural networks, unsupervised learning, supervised learning

About these checks

Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.

Statistics and arithmetic are recomputed, not judged.

Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.

What counts as a claim ›How we measure confidence ›How we check this ›

Are you an author, or do you see a problem with this page? Request a review

How to cite

Cite CoChat for the checks on this page, and the paper for its own findings.

This page
CoChat. (2026). Citation check: Scalable training of artificial neural networks with adaptive sparse connectivity inspired by network science. https://app.cochat.ai/papers/10.1038/s41467-018-04316-3
The paper
Mocanu, D. C., Mocanu, E., Stone, P., Nguyen, P. H., Gibescu, M., & Liotta, A. (2018). Scalable training of artificial neural networks with adaptive sparse connectivity inspired by network science. Nature Communications. https://doi.org/10.1038/s41467-018-04316-3

Data sources

Bibliographic data: Semantic Scholar, OpenAlex, Crossref.

Contains information from the Semantic Scholar Academic Graph Datasets, made available under the ODC Attribution License (ODC-BY 1.0).

Semantic Scholar’s data is described in: Kinney, R., et al. (2023). The Semantic Scholar Open Data Platform. arXiv:2301.10140.