Filters








2,692 Hits in 7.0 sec

Fundamental limits of Bayesian inference: order parameters and phase transitions for road tracking

A.L. Yuille, J.M. Coughlan
2000 IEEE Transactions on Pattern Analysis and Machine Intelligence  
We demonstrate that there is a phase transition at a critical value of the order parameter K -below this phase transition it is impossible to detect the road by any algorithm.  ...  There is a growing interest in formulating vision problems in terms of Bayesian inference and, in particular, the maximimum a posteriori (MAP) estimator.  ...  We would also like to thank Dan Snow and Scott Konishi for helpful discussions as the work was progressing and Davi Geiger for providing useful stimulation.  ... 
doi:10.1109/34.825754 fatcat:ylvg42vldzg2hnpyj5tzrno2hi

Statistical physics of linear and bilinear inference problems [article]

Christophe Schülke
2016 arXiv   pre-print
Third, a theoretical analysis of matrix compressed sensing and of instabilities in Bayesian bilinear inference algorithms.  ...  The aim of this thesis is to propose efficient algorithms for this class of problems and to perform their theoretical analysis.  ...  The main results are the definition of a new set of order parameters, the existence of multiple phase transitions and a study of several real networks.  ... 
arXiv:1607.00675v1 fatcat:jw4fq7svxvfpvgsm4rob6pa2fm

Occam's Razor Applied to Network Topology Inference

Dimitri Marinakis, Gregory Dudek
2008 IEEE Transactions on robotics  
We present a method for inferring the topology of a sensor network given nondiscriminating observations of activity in the monitored region.  ...  Our approach employs a two-level reasoning system made up of a stochastic expectation maximization algorithm and a higher level search strategy employing the principle of Occam's Razor to look for the  ...  The first level is made up of our fundamental topology inference algorithm that takes the sensor observations and environmental assumptions as inputs and returns the network parameters.  ... 
doi:10.1109/tro.2008.918048 fatcat:wj3o3xtp5zf4fcqr6fot6v6yz4

Brain-Inspired Hardware Solutions for Inference in Bayesian Networks

Leila Bagheriye, Johan Kwisthout
2021 Frontiers in Neuroscience  
The implementation of inference (i.e., computing posterior probabilities) in Bayesian networks using a conventional computing paradigm turns out to be inefficient in terms of energy, time, and space, due  ...  A departure from conventional computing systems to make use of the high parallelism of Bayesian inference has attracted recent attention, particularly in the hardware implementation of Bayesian networks  ...  by five orders of magnitude for Bayesian inference.  ... 
doi:10.3389/fnins.2021.728086 pmid:34924925 pmcid:PMC8677599 fatcat:tihogzl6tfbpjdybwpggllwd5u

Parameter Inference for Computational Cognitive Models with Approximate Bayesian Computation

Antti Kangasrääsiö, Jussi P. P. Jokinen, Antti Oulasvirta, Andrew Howes, Samuel Kaski
2019 Cognitive Science  
The results contrast the efficiency and informativeness of the methods. A key advantage of the Bayesian methods is the ability to estimate the uncertainty of fitted parameter values.  ...  In this article, we investigate the capability and role of modern fitting methods-including Bayesian optimization and approximate Bayesian computation-and contrast them to some more commonly used methods  ...  The inference for the ACT-R model was not parallelized, and the inference was limited to 3 GB of memory. The RL model was parallelized using 20 worker processes, each limited to 6 GB of memory.  ... 
doi:10.1111/cogs.12738 pmid:31204797 pmcid:PMC6593436 fatcat:dztycr4gvfey3aaxq3kycwbt44

PID Control as a Process of Active Inference with Linear Generative Models

Manuel Baltieri, Christopher Buckley
2019 Entropy  
In particular, the free energy principle and active inference are increasingly popular theories of cognitive functions that claim to offer a unified understanding of life and cognition within a general  ...  This more general interpretation also provides a new perspective on traditional problems of PID controllers such as parameter tuning as well as the need to balance performances and robustness conditions  ...  a process consistent with a Bayesian inference scheme.  ... 
doi:10.3390/e21030257 pmid:33266972 fatcat:pq5xwm6yfrfannofroij44z5ma

The Path Inference Filter: Model-Based Low-Latency Map Matching of Probe Vehicle Data [chapter]

Timothy Hunter, Pieter Abbeel, Alexandre M. Bayen
2013 Springer Tracts in Advanced Robotics  
We introduce a new class of algorithms, which are altogether called the path inference filter (PIF), that maps GPS data in real time, for a variety of tradeoffs and scenarios and with a high throughput  ...  We consider the problem of reconstructing vehicle trajectories from sparse sequences of GPS points, for which the sampling interval is between 1 s and 2 min.  ...  Samarayanake for the discussions that have been instrumental in writing this paper; and W. Hoburg and K. Kroetz for their thorough and insightful comments on the draft.  ... 
doi:10.1007/978-3-642-36279-8_36 fatcat:75gj6fmsvrbpfk7tpe7lhbariq

To design interfaces for exploratory data analysis, we need theories of graphical inference [article]

Jessica Hullman, Andrew Gelman
2021 arXiv   pre-print
Implications of Bayesian and other theories of graphical inference should be tested against outcomes of interactive analysis by people to drive theory development.  ...  But design philosophies that emphasize exploration over other phases of analysis risk confusing a need for flexibility with a conclusion that exploratory visual analysis is inherently model-free and cannot  ...  Office of Naval Research for grant N000141912204, the Institute for Education Sciences for grant R305D190048, and Alex Kale, Jeffrey Heer, Hanspeter Pfister, Matthew Kay, Jennie Rogers, and the anonymous  ... 
arXiv:2104.02015v2 fatcat:kqfnd7pz6vhuvp5p3peaxlvtli

The Path Inference Filter: Model-Based Low-Latency Map Matching of Probe Vehicle Data

Timothy Hunter, Pieter Abbeel, Alexandre Bayen
2014 IEEE transactions on intelligent transportation systems (Print)  
We introduce a new class of algorithms, which are altogether called the path inference filter (PIF), that maps GPS data in real time, for a variety of tradeoffs and scenarios and with a high throughput  ...  We consider the problem of reconstructing vehicle trajectories from sparse sequences of GPS points, for which the sampling interval is between 1 s and 2 min.  ...  Samarayanake for the discussions that have been instrumental in writing this paper; and W. Hoburg and K. Kroetz for their thorough and insightful comments on the draft.  ... 
doi:10.1109/tits.2013.2282352 fatcat:yosbyscnxbchli5lvacsulgqn4

Deeply Felt Affect: The Emergence of Valence in Deep Active Inference

Casper Hesp, Ryan Smith, Thomas Parr, Micah Allen, Karl J. Friston, Maxwell J. D. Ramstead
2020 Neural Computation  
Using deep active inference, a hierarchical inference scheme that rests on inverting a model of how sensory data are generated, we develop a principled Bayesian model of emotional valence.  ...  This index of subjective fitness can be estimated within any environment and exploits the domain generality of second-order beliefs (beliefs about beliefs).  ...  each comprising an initial phase of negative valence (anxiety; quarters 1 and 3), followed by a phase of positive valence (confidence; quarters 2 and 4).  ... 
doi:10.1162/neco_a_01341 pmid:33253028 pmcid:PMC8594962 fatcat:zy32lkkdmna5ln6n6gajqbfjgq

Active Inference in Modeling Conflict [article]

Scott David, Richard J. Cordes, Daniel A. Friedman
2021 Zenodo  
Insights and implications from qualitative use are used as a foundation for offering recommendations for future research and social systems design.  ...  This formalization, the Active Inference Conflict (AIC) model, situates conflict in terms of a multiscale process of communication, trust, and relationship management enacted by interacting entities.  ...  other drivers in order to share the road.  ... 
doi:10.5281/zenodo.5759800 fatcat:pmo7isoig5a3joav2z2ydyh3v4

COSMIC BIRTH: Efficient Bayesian Inference of the Evolving Cosmic Web from Galaxy Surveys [article]

Francisco-Shu Kitaura, Metin Ata, Sergio A. Rodriguez-Torres, Monica Hernandez-Sanchez, A. Balaguera-Antolinez, Gustavo Yepes
2020 arXiv   pre-print
We present COSMIC BIRTH: COSMological Initial Conditions from Bayesian Inference Reconstructions with THeoretical models: an algorithm to reconstruct the primordial and evolved cosmic density fields from  ...  Novel key ingredients to this approach are a higher-order Hamiltonian sampling technique and a non-diagonal Hamiltonian mass-matrix.  ...  ACKNOWLEDGMENTS We thank the referee for the careful revision of the manuscript. The authors thank Raúl E. Angulo, Ginevra Favole, Mariana Vargas-Magaña, and Cheng Zhao for discussions.  ... 
arXiv:1911.00284v2 fatcat:czbsgmmnobgubkkgjnxzdnudea

Inference, prediction and optimization of non-pharmaceutical interventions using compartment models: the PyRoss library [article]

R. Adhikari, Austen Bolitho, Fernando Caballero, Michael E. Cates, Jakub Dolezal, Timothy Ekeh, Jules Guioth, Robert L. Jack, Julian Kappler, Lukas Kikuchi, Hideki Kobayashi, Yuting I. Li (+6 others)
2020 arXiv   pre-print
PyRoss is an open-source Python library that offers an integrated platform for inference, prediction and optimisation of NPIs in age- and contact-structured epidemiological compartment models.  ...  PyRoss allows fully Bayesian forecasts of the impact of idealized NPIs by convolving uncertainties arising from epidemiological data, model choice, parameters, and intrinsic stochasticity.  ...  We thank the code review team of RAMP's Rapid Review Group at Oxford for their scrutiny of the PyRoss library and for their suggestions for improvement; RAMP's Red Team at Edinburgh further code review  ... 
arXiv:2005.09625v1 fatcat:ogkpp4hdu5hj3p4syvzkddgtp4

Great Expectations: Unsupervised Inference of Suspense, Surprise and Salience in Storytelling [article]

David Wilmot
2022 arXiv   pre-print
Extensions add memory and external knowledge from story plots and from Wikipedia to infer salience on novels such as Great Expectations and plays such as Macbeth.  ...  Stories interest us not because they are a sequence of mundane and predictable events but because they have drama and tension.  ...  c) Models are limited in the amount of knowledge they can represent by the number of parameters.  ... 
arXiv:2206.09708v1 fatcat:k4oefywyxvgn5gdtedyvr5mbpi

Counterfactual reasoning about intent for interactive navigation in dynamic environments

Alejandro Bordallo, Fabio Previtali, Nantas Nardelli, Subramanian Ramamoorthy
2015 2015 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS)  
This inference framework is coupled with a novel distributed visual tracking method that provides reliable and robust models for the current belief-state of the monitored environment.  ...  By using a light-weight motion model, we achieve efficient iterative planning for fluid motion when avoiding pedestrians, in parallel with goal inference for longer range movement prediction.  ...  This is the most difficult and fundamental step for any tracking algorithm. In our approach, we consider as features for data association the direction, the velocity and the position of the objects.  ... 
doi:10.1109/iros.2015.7353783 dblp:conf/iros/BordalloPNR15 fatcat:jfjpe3w7q5fdxfagfl2avneoha
« Previous Showing results 1 — 15 out of 2,692 results