Filters








3,048 Hits in 6.4 sec

Linear-Time Sequence Classification using Restricted Boltzmann Machines [article]

Son N. Tran, Srikanth Cherla, Artur Garcez, Tillman Weyde
2018 arXiv   pre-print
Classification of sequence data is the topic of interest for dynamic Bayesian models and Recurrent Neural Networks (RNNs).  ...  Also, the experimental results on optical character recognition, part-of-speech tagging and text chunking demonstrate that our model is comparable to recurrent neural networks with complex memory gates  ...  It indicates a consistent improvement in best-case performance from the n-gram models, the non-recurrent neural networks, and then the recurrent neural network models, with the SCRBM outperforming all  ... 
arXiv:1710.02245v3 fatcat:jq6cu4ztt5gl7l66gf6cws3zna

Tracking slow modulations in synaptic gain using dynamic causal modelling: Validation in epilepsy

Margarita Papadopoulou, Marco Leite, Pieter van Mierlo, Kristl Vonck, Louis Lemieux, Karl Friston, Daniele Marinazzo
2015 NeuroImage  
Bayesian model selection was used to identify the intrinsic (within-source) and extrinsic (between-source) connectivity.  ...  Our key finding was that intrinsic synaptic changes were sufficient to explain seizure onset, where these changes showed dissociable time courses over several seconds.  ...  Dynamic causal modelling of cross spectral density can be implemented using the DCM Toolbox.  ... 
doi:10.1016/j.neuroimage.2014.12.007 pmid:25498428 pmcid:PMC4306529 fatcat:q7yss2ng2vhw7egfaoo37tdh44

OutbreakFlow: Model-based Bayesian inference of disease outbreak dynamics with invertible neural networks and its application to the COVID-19 pandemics in Germany [article]

Stefan T. Radev, Frederik Graw, Simiao Chen, Nico T. Mutters, Vanessa M. Eichel, Till Bärnighausen, Ullrich Köthe
2021 arXiv   pre-print
In this work, we address this problem with a novel combination of epidemiological modeling with specialized neural networks.  ...  In the subsequent inference phase, the trained neural network processes the observed data of an actual outbreak and infers the parameters of the model in order to realistically reproduce the observed dynamics  ...  Our neural architecture comprises three sub-networks: (i) a convolutional filtering network performing noise reduction and feature extraction on the raw observational data; (ii) a recurrent summary network  ... 
arXiv:2010.00300v4 fatcat:jqjs5cgdwbe3tnsfttrqtutiaq

Amortized Bayesian model comparison with evidential deep learning [article]

Stefan T. Radev, Marco D'Alessandro, Ulf K. Mertens, Andreas Voss, Ullrich Köthe, Paul-Christian Bürkner
2021 arXiv   pre-print
We demonstrate the utility of our method on toy examples and simulated data from non-trivial models from cognitive science and single-cell neuroscience.  ...  The Bayesian probabilistic framework offers a principled way to perform model comparison and extract useful metrics for guiding decisions.  ...  We also thank David Izydorczyk and Mattia Sensi for reading the paper and providing constructive feedback.  ... 
arXiv:2004.10629v4 fatcat:gpsjtnxm4bfftowftsdvqdjs7q

Neural Integration of Continuous Dynamics [article]

Margaret Trautner, Sai Ravela
2019 arXiv   pre-print
Modeled as constant-sized recurrent networks embedding a continuous neural differential equation, they achieve fully neural temporal output.  ...  Using the polynomial class of dynamical systems, we demonstrate the equivalence of neural and numerical integration.  ...  Support from ONR grant N00014-19-1-2273, the MIT Environmental Solutions Initiative, the Maryanne and John Montrym Fund, and the MIT Lincoln Laboratory are gratefully acknowledged.  ... 
arXiv:1911.10309v1 fatcat:ewjuh4hewbd23odgarsnev4zhi

Recognizing recurrent neural networks (rRNN): Bayesian inference for recurrent neural networks

Sebastian Bitzer, Stefan J. Kiebel
2012 Biological cybernetics  
We suggest that the Bayesian inversion of recurrent neural networks may be useful both as a model of brain function and as a machine learning tool.  ...  Recurrent neural networks (RNNs) are widely used in computational neuroscience and machine learning applications.  ...  Acknowledgments We thank both anonymous reviewers for the helpful and constructive comments on a previous version of this manuscript.  ... 
doi:10.1007/s00422-012-0490-x pmid:22581026 fatcat:y3prg6rhfjg2bivyndcydzmsoq

OutbreakFlow: Model-based Bayesian inference of disease outbreak dynamics with invertible neural networks and its application to the COVID-19 pandemics in Germany

Stefan T. Radev, Frederik Graw, Simiao Chen, Nico T. Mutters, Vanessa M. Eichel, Till Bärnighausen, Ullrich Köthe, Mark M. Tanaka
2021 PLoS Computational Biology  
In this work, we address this problem with a novel combination of epidemiological modeling with specialized neural networks.  ...  In the subsequent inference phase, the trained neural network processes the observed data of an actual outbreak and infers the parameters of the model in order to realistically reproduce the observed dynamics  ...  Our neural architecture comprises three sub-networks: (i) a convolutional filtering network performing noise reduction and feature extraction on the raw observational data; (ii) a recurrent summary network  ... 
doi:10.1371/journal.pcbi.1009472 pmid:34695111 pmcid:PMC8584772 fatcat:tz6ei3zd6rfd7g5ir6a7uvfy3a

Survey of Cryptocurrency Volatility Prediction Literature Using Artificial Neural Networks

Sina E. Charandabi, Kamyar Kamyar
2022 Business and Economic Research  
Recently-developed literature that attempt to predict volatilities of cryptocurrency valuations through creation of hybrid artificial neural network models are then discussed.  ...  For the major part of the paper, we delve into details of multiple hybrid artificial neural networks that were thoroughly implemented to predict cryptocurrency volatilities.  ...  The model is established on the quadratic variance of the theory of arbitrage-free price processes in time series, and it uses connections among realized volatility and matrix of conditional covariance  ... 
doi:10.5296/ber.v12i1.19301 fatcat:pst3igfhivdc5pi2ln4ubcdyae

Towards Automated Satellite Conjunction Management with Bayesian Deep Learning [article]

Francesco Pinto, Giacomo Acciarini, Sascha Metz, Sarah Boufelja, Sylvester Kaczmarek, Klaus Merz, José A. Martinez-Heras, Francesca Letizia, Christopher Bridges, Atılım Güneş Baydin
2020 arXiv   pre-print
We introduce a Bayesian deep learning approach to this problem, and develop recurrent neural network architectures (LSTMs) that work with time series of conjunction data messages (CDMs), a standard data  ...  We show that our method can be used to model all CDM features simultaneously, including the time of arrival of future CDMs, providing predictions of conjunction event evolution with associated uncertainties  ...  We would like to thank Dario Izzo and Moriba Jah for sharing their technical expertise and James Parr, Jodie Hughes, Leo Silverberg, Alessandro Donati for their support.  ... 
arXiv:2012.12450v1 fatcat:dadqsub7uvevlfx32ce645mziu

Connectivity Inference from Neural Recording Data: Challenges, Mathematical Bases and Research Directions [article]

Ildefons Magrans de Abril, Junichiro Yoshimoto, Kenji Doya
2017 arXiv   pre-print
We then review connectivity inference methods based on two major mathematical foundations, namely, descriptive model-free approaches and generative model-based approaches.  ...  We first identify biophysical and technical challenges in connectivity inference along the data processing pipeline.  ...  and internal funding from the Okinawa Institute of Science and Technology Graduate University.  ... 
arXiv:1708.01888v2 fatcat:fezbmzuzenac7mqcnqhq5sveye

The graphical brain: Belief propagation and active inference

Karl J. Friston, Thomas Parr, Bert de Vries
2017 Network Neuroscience  
For example, Bayesian model averaging and comparison, which link discrete and continuous states, may be implemented in thalamocortical loops.  ...  To accommodate mixed generative models (of discrete and continuous states), one also has to consider link nodes or factors that enable discrete and continuous representations to talk to each other.  ...  Note the formal similarity between the Bayesian network and the Forney factor graph; however, also note the differences.  ... 
doi:10.1162/netn_a_00018 pmid:29417960 pmcid:PMC5798592 fatcat:ew5x2cczwvarfeedfz5s6ldivm

FORECASTING FOREIGN EXCHANGE RATES WITH ARTIFICIAL NEURAL NETWORKS: A REVIEW

WEI HUANG, K. K. LAI, Y. NAKAMORI, SHOUYANG WANG
2004 International Journal of Information Technology and Decision Making  
Several design factors significantly impact the accuracy of neural network forecasts. These factors include the selection of input variables, preparing data, and network architecture.  ...  We also describe the integration of ANNs with other methods and report the comparison between performances of ANNs and those of other forecasting methods, and finding mixed results.  ...  Acknowledgement This project is supported by NSFC, CAS and the City University of Hong Kong.  ... 
doi:10.1142/s0219622004000969 fatcat:5woran6t6veidh373r6ibivfmm

Self-Supervised Inference in State-Space Models [article]

David Ruhe, Patrick Forré
2022 arXiv   pre-print
Without parameterizing a generative model, we apply Bayesian update formulas using a local linearity approximation parameterized by neural networks.  ...  Usage of such domain knowledge is reflected in excellent results (despite our model's simplicity) on the chaotic Lorenz system compared to fully supervised and variational inference methods.  ...  The model obtained by parameterizing p(x k | y <k ) directly (eq. ( 12 )) is referred to as the recurrent filter or recurrent smoother, as it only employs recurrent neural networks (and no Bayesian recursion  ... 
arXiv:2107.13349v3 fatcat:rp2ionq6kbbblh6m563pg6mn4y

Learning Scalable Deep Kernels with Recurrent Structure

Maruan Al-Shedivat, Andrew Gordon Wilson, Yunus Saatchi, Zhiting Hu, Eric P Xing
2017 Journal of machine learning research  
The resulting model, GP-LSTM, fully encapsulates the inductive biases of long short-term memory (LSTM) recurrent networks, while retaining the non-parametric probabilistic advantages of Gaussian processes  ...  Many applications in speech, robotics, finance, and biology deal with sequential data, where ordering matters and recurrent structures are common.  ...  This work was supported in part by NIH R01GM114311, AFRL/DARPA FA87501220324, and NSF IIS-1563887.  ... 
pmid:30662374 pmcid:PMC6334642 fatcat:oloxsfrnvvh53kouiredghe3tm

The time dimension of neural network models

Richard Rohwer
1994 ACM SIGART Bulletin  
This review attempts to provide an insightful perspective on the role of time within neural network models and the use of neural networks for problems involving time.  ...  The most commonly used neural network models are de ned and explained giving mention to important technical issues but avoiding great detail.  ...  It is interesting to note a formal similarity between recurrent networks and TDNNs.  ... 
doi:10.1145/181911.181917 fatcat:qk5nunmmrzcohncrnm4k7wd3uq
« Previous Showing results 1 — 15 out of 3,048 results