The DOI matches its Crossref record; 117 of the 122 references it cites were found in a registry or an index.
Practical Recommendations for Gradient-Based Training of Deep Architectures
Open accessAbstract
Learning algorithms related to artificial neural networks and in particular for Deep Learning may seem to involve many bells and whistles, called hyper-parameters. This chapter is meant as a practical guide with recommendations for some of the most commonly used hyperparameters, in particular in the context of learning algorithms based on back-propagated gradient and gradient-based optimization. It also discusses how to deal with the fact that more interesting results can be obtained when allowing one to adjust many hyper-parameters. Overall, it describes elements of the practice used to successfully and efficiently train and debug large-scale and often deep multi-layer neural networks. It closes with open questions about the training difficulties observed with deeper architectures.
Overview
Integrity signals
Registry
Cites
Statistics
Verification
Claims the paper makes
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?
In the citing paper
“The choice of those parameters has been shown to have a high impact on the final classification performances ([13,26] for HAR using wearable sensors).”
From Comparison of Feature Learning Methods for Human Activity Recognition Using Wearable Sensors (2018) · cites this paper as [26]
In the citing paper
“using the same number of neurons in every layer, rather than a sequential decrease or increase of them, seems to be a better choice.”
From A review of multimodal deep learning methods for genomic-enabled prediction in plant breeding (2024) · cites this paper as Bengio 2012
In the citing paper
“the same paper claims that such advice is based on experimental results and every data set could require a specific choice for such hyperparameters.”
From A review of multimodal deep learning methods for genomic-enabled prediction in plant breeding (2024) · cites this paper as Bengio 2012
In the citing paper
“in the study by Bengio 2012, 2012 it is suggested to include several nodes large enough in every layer”
From A review of multimodal deep learning methods for genomic-enabled prediction in plant breeding (2024) · cites this paper as Bengio 2012
Citations
Cites
Gradient-based learning applied to document recognition
Lecun, Y., Bottou, L., Bengio, Y., & Haffner, P. (1998) · Proceedings of the IEEE · doi:10.1109/5.726791 · Cited by 58,805 · cited 9×
“LeCun et al. (1998a) also propose a simple method in order to estimate the largest eigenvalue of the Hessian (which would be the optimal learning rate).”
BackgroundMethodInfluentialSelf-citation(2008) · Journal of Machine Learning Research · Cited by 33,051
Learning representations by back-propagating errors
Rumelhart, D. E., Hinton, G. E., & Williams, R. J. (1986) · Nature · doi:10.1038/323533a0 · Cited by 31,892 · cited 1×
“This chapter assumes that the reader already un-derstands the standard algorithms for training supervised multi-layer neural networks, with the loss gradient computed thanks to the back-propagation algorithm (Rumelhart et al. , 1986).”
MethodA fast learning algorithm for deep belief nets
Hinton, G. E., Osindero, S., & Teh, Y.-W. (2006) · Neural Computation · doi:10.1162/neco.2006.18.7.1527 · Cited by 16,336 · cited 13×
“Note that some learning algorithms (in particular unsupervised learning algorithms such as algorithms for training RBMs by approximate maximum likelihood) are problematic in this respect because we cannot directly measure the quantity that is to be optimized (e.g. the likelihood) because it is intractable.”
BackgroundMethodInfluentialBreiman, L. (1996) · Machine Learning · doi:10.1007/bf00058655 · Cited by 16,325 · cited 1×
“Another way to exploit computing power to push performance a bit is model averaging, as in Bagging (Breiman, 1994) and Bayesian methods.”
Method(2010) · Cited by 14,134
Self-citation(2010) · International Conference on Machine Learning · Cited by 13,995
A Global Geometric Framework for Nonlinear Dimensionality Reduction
Tenenbaum, J. B., Silva, V. D., & Langford, J. C. (2000) · Science · doi:10.1126/science.290.5500.2319 · Cited by 13,585 · cited 2×
“D) representation, e.g., with the t-SNE (van der Maaten and Hinton, 2008) or Isomap (Tenenbaum et al. , 2000) algorithms, and then plot the evolution of this function during training, or the population of such trajectories for different initializations.”
BackgroundMethodIndexing by latent semantic analysis
Deerwester, S., Dumais, S. T., Furnas, G. W., Landauer, T. K., & Harshman, R. (1990) · Journal of the American Society for Information Science · doi:10.1002/(sici)1097-4571(199009)41:6<391::aid-asi1>3.0.co;2-9 · Cited by 12,751
Backpropagation Applied to Handwritten Zip Code Recognition
LeCun, Y., Boser, B., Denker, J. S., Henderson, D., Howard, R. E., Hubbard, W., & Jackel, L. D. (1989) · Neural Computation · doi:10.1162/neco.1989.1.4.541 · Cited by 11,891 · cited 1×
“There are also many architectural choices that are relevant in the case of convolutional architectures (e.g. for modeling images, time-series or sound) (LeCun et al. , 1989, 1998b; Le et al. , 2010) in which hidden units have local receptive fields.”
Background
Research works printed but not linked: 10
- Bach, F., Moulines, E.: Non-asymptotic analysis of stochastic approximation algorithms. In: NIPS 2011 (2011)
- Bengio, Y., Ducharme, R., Vincent, P., Jauvin, C.: A neural probabilistic language model. JMLR 3, 1137–1155 (2003)
- Bottou, L., LeCun, Y.: Large-scale on-line learning. In: NIPS 2003 (2004)
- Caruana, R.: Multitask connectionist learning. In: Proceedings of the 1993 Connectionist Models Summer School, pp. 372–379 (1993)
- Collobert, R., Bengio, S.: Links between perceptrons, MLPs and SVMs. In: International Conference on Machine Learning, ICML (2004b) · DOI · Paper page
- Dauphin, Y., Glorot, X., Bengio, Y.: Sampled reconstruction for large-scale learning of embeddings. In: Proc. ICML 2011 (2011)
- Le Roux, N., Manzagol, P.-A., Bengio, Y.: Topmoumoute online natural gradient algorithm. In: NIPS 2007 (2008)
- LeCun, Y.: Modèles connexionistes de l’apprentissage. Ph.D. thesis, Université de Paris VI (1987)
- Mesnil, G., Dauphin, Y., Glorot, X., Rifai, S., Bengio, Y., Goodfellow, I., Lavoie, E., Muller, X., Desjardins, G., Warde-Farley, D., Vincent, P., Courville, A., Bergstra, J.: Unsupervised and transfer learning challenge: a deep learning approach. In: Proc. Unsupervised and Transfer Learning, JMLR W&CP, vol. 7 (2011)
- Ranzato, M., Boureau, Y.-L., LeCun, Y.: Sparse feature learning for deep belief networks. In: Platt, J., Koller, D., Singer, Y., Roweis, S. (eds.) Advances in Neural Information Processing Systems (NIPS 2007), vol. 20, pp. 1185–1192. MIT Press, Cambridge (2008a)
Cited by
A survey on deep learning in medical image analysis
Litjens, G., Kooi, T., Bejnordi, B. E., Setio, A. A. A., Ciompi, F., Ghafoorian, M., van der Laak, J. A. W. M., van Ginneken, B., & Sánchez, C. I. (2017) · Medical Image Analysis · doi:10.1016/j.media.2017.07.005 · Cited by 15,332
“Most researchers fall back to an intuition-based random search (Bergstra and Bengio, 2012), which often seems to work well enough.”
BackgroundInfluentialRepresentation Learning: A Review and New Perspectives
Bengio, Y., Courville, A., & Vincent, P. (2013) · IEEE Transactions on Pattern Analysis and Machine Intelligence · doi:10.1109/tpami.2013.50 · Cited by 13,001
“more modern compendium of good training practice, particularly adapted to training RBMs, is provided in Hinton (2010), while a similar guide oriented more towards deep neural networks can be found in Bengio (2013), both of which are part of a novel version of the above book. Recent work on automating hyper-parameter s…”
BackgroundInfluentialSelf-citationBrain tumor segmentation with Deep Neural Networks
Havaei, M., Davy, A., Warde-Farley, D., Biard, A., Courville, A., Bengio, Y., Pal, C., Jodoin, P.-M., & Larochelle, H. (2017) · Medical Image Analysis · doi:10.1016/j.media.2016.05.004 · Cited by 3,278
“Hyper-parameters were tuned using grid search and cross-validation on a validation set (see Bengio [6]).”
MethodSelf-citationDeep Learning‐Based Crack Damage Detection Using Convolutional Neural Networks
Cha, Y., Choi, W., & Büyüköztürk, O. (2017) · Computer-Aided Civil and Infrastructure Engineering · doi:10.1111/mice.12263 · Cited by 3,295
“However, several useful articles can be found from Bottou (2012), LeCun et al. (2012), and Bengio (2012).”
BackgroundMethodDeep learning for healthcare: review, opportunities and challenges
Miotto, R., Wang, F., Wang, S., Jiang, X., & Dudley, J. T. (2018) · Briefings in Bioinformatics · doi:10.1093/bib/bbx044 · Cited by 3,076
Metadata
- Authors
- Bengio, Y.
- Affiliations
- Université de Montréal (CA)
- Corresponding author
- Bengio, Yoshua
- ORCID
- Bengio, Yoshua
- Venue
- Lecture Notes in Computer Science
- Publisher
- Springer Berlin Heidelberg
- Published
- 2012
- Pages
- 437-478
- ISSN
- 0302-9743 (ISSN-L)
- 1611-3349
- OpenAlex
- W1806891645
- Cited by
- 2,259
- References
- 122
- Citations by year
- 2026: 95 · 2025: 186 · 2024: 184 · 2023: 195 · 2022: 253 · 2021: 273 · 2020: 278 · 2019: 231 · 2018: 200 · 2017: 158 · 2016: 105 · 2015: 55 · 2014: 26 · 2013: 13 · 2012: 7
- FWCI
- 24.71
- Type
- conference paper
- Language
- en
- Open access
- green
- arxiv.org
- Locations
- Lecture notes in computer science (publishedVersion)
- arXiv (Cornell University) (submittedVersion)
- arXiv (Cornell University)
- Funders
- Canadian Institute for Advanced Research
- Canada Research Chairs
- Fonds Québécois de la Recherche sur la Nature et les Technologies
- Natural Sciences and Engineering Research Council of Canada
- Topics
- Physical Sciences › Computer Science › Artificial Intelligence › Neural Networks and Applications
- Physical Sciences › Computer Science › Artificial Intelligence › Stochastic Gradient Optimization Techniques
- Physical Sciences › Computer Science › Artificial Intelligence › Machine Learning and Data Classification
- Keywords
- hyperparameter optimization, gradient-based optimization, backpropagation, deep neural network, training recommendations, multilayer neural networks
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: Practical Recommendations for Gradient-Based Training of Deep Architectures. https://app.cochat.ai/papers/10.1007/978-3-642-35289-8_26
- The paper
- Bengio, Y. (2012). Practical Recommendations for Gradient-Based Training of Deep Architectures. Lecture Notes in Computer Science. https://doi.org/10.1007/978-3-642-35289-8_26
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.