A Fixed Size Storage O(n3) Time Complexity Learning Algorithm for Fully Recurrent Continually Running Networks
The DOI matches its Crossref record.
Abstract
The real-time recurrent learning (RTRL) algorithm (Robinson and Fallside 1987; Williams and Zipser 1989) requires O(n 4) computations per time step, where n is the number of noninput units. I describe a method suited for on-line learning that computes exactly the same gradient and requires fixed-size storage of the same order but has an average time complexity per time step of O(n 3).
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.
Integrity signals
Registry
Cites
Statistics
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.
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
Cited by
Hochreiter, S., & Schmidhuber, J. (1997) · Neural Computation · doi:10.1162/neco.1997.9.8.1735 · Cited by 102,318
“BPTT: Same architecture as the one trained by RTRL. CH: Both net architectures like RTRL’s, but one has an additional output for predicting the hidden unit of the other one (see Schmidhuber, 1992b, for details).”
BackgroundMethodInfluentialSelf-citationDeep learning in neural networks: an overview
Schmidhuber, J. (2015) · Neural Networks · doi:10.1016/j.neunet.2014.09.003 · Cited by 18,570
Self-citationRecurrent neural networks and robust time series prediction
Connor, J. T., Martin, R. D., & Atlas, L. E. (1994) · IEEE Transactions on Neural Networks · doi:10.1109/72.279188 · Cited by 1,316
Gradient calculations for dynamic recurrent neural networks: a survey
Pearlmutter, B. A. (1995) · IEEE Transactions on Neural Networks · doi:10.1109/72.410363 · Cited by 582
New results on recurrent network training: unifying the algorithms and accelerating convergence
Atiya, A. F., & Parlos, A. G. (2000) · IEEE Transactions on Neural Networks · doi:10.1109/72.846741 · Cited by 521
Learning Complex, Extended Sequences Using the Principle of History Compression
Schmidhuber, J. (1992) · Neural Computation · doi:10.1162/neco.1992.4.2.234 · Cited by 398
Self-citationLearning to Control Fast-Weight Memories: An Alternative to Dynamic Recurrent Networks
Schmidhuber, J. (1992) · Neural Computation · doi:10.1162/neco.1992.4.1.131 · Cited by 357
Self-citation(2000) · CrossRef Listing of Deleted DOIs · doi:10.1162/153244303768966139 · Cited by 289
Developmental robotics, optimal artificial curiosity, creativity, music, and the fine arts
Schmidhuber, J. (2006) · Connection Science · doi:10.1080/09540090600768658 · Cited by 268
Self-citationTraining Recurrent Networks by Evolino
Schmidhuber, J., Wierstra, D., Gagliolo, M., & Gomez, F. (2007) · Neural Computation · doi:10.1162/neco.2007.19.3.757 · Cited by 248
“Applied to the LSTM architecture, Evolino can solve tasks that ESNs [15] cannot, and achieves higher accuracy in certain continuous function generation tasks than conventional gradient descent RNNs, including gradient-based LSTM (henceforth called G-LSTM).”
BackgroundInfluentialSelf-citationAn analog VLSI recurrent neural network learning a continuous-time trajectory
Cauwenberghs, G. (1996) · IEEE Transactions on Neural Networks · doi:10.1109/72.485671 · Cited by 127
Schmidhuber, J. (1993) · ICANN ’93 · doi:10.1007/978-1-4471-2063-6_110 · Cited by 34
“(1)It should be noted that there is a fixed-size storage algorithm (a hybrid between BPTT and RTRL) with Rtime = O(m(2)) (better than with RTRL, worse than with BPTT) and Rspace = O(m (2)) (much better than with BPTT, same as with RTRL) ([7][4]).”
BackgroundSelf-citation
Metadata
- Authors
- Schmidhuber, J.
- Venue
- Neural Computation
- Publisher
- MIT Press - Journals
- Published
- 1992-03
- Volume
- 4
- Issue
- 2
- Pages
- 243-248
- ISSN
- 0899-7667 (ISSN-L)
- 1530-888X
- OpenAlex
- W2114471683
- Cited by
- 110
- References
- 0
- FWCI
- 6.26
- Type
- journal article
- Language
- en
- Open access
- closed
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
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.