Filters








118,324 Hits in 3.6 sec

A Simple Stochastic Algorithm for Structural Features Learning [chapter]

Jan Mačák, Ondřej Drbohlav
2015 Lecture Notes in Computer Science  
A conceptually very simple unsupervised algorithm for learning structure in the form of a hierarchical probabilistic model is described in this paper.  ...  All these learned models constitute a rich set of independent structure elements of variable complexity that can be used as features in various recognition tasks.  ...  Summary In this paper, a very simple stochastic algorithm for unsupervised joint learning of structure and parameters is described.  ... 
doi:10.1007/978-3-319-16634-6_4 fatcat:rchywuurgzbufnkyi2sxb4q7nu

Learning a Distance Metric from a Network

Blake Shaw, Bert Huang, Tony Jebara
2011 Neural Information Processing Systems  
To better model and understand these networks, we present structure preserving metric learning (SPML), an algorithm for learning a Mahalanobis distance metric from a network such that the learned distances  ...  Like the graph embedding algorithm structure preserving embedding, SPML learns a metric which is structure preserving, meaning a connectivity algorithm such as k-nearest neighbors will yield the correct  ...  For SPML, we learn a full M via Algorithm 3. For each person, we construct a sparse feature vector where there is one feature corresponding to every possible dorm, major, etc. for each feature type.  ... 
dblp:conf/nips/ShawHJ11 fatcat:nqnmjqrniveulgl7zausxyp2ui

A critical evaluation of stochastic algorithms for convex optimization

Simon Wiesler, Alexander Richard, Ralf Schluter, Hermann Ney
2013 2013 IEEE International Conference on Acoustics, Speech and Signal Processing  
We obtained slight improvements by using a stochastic second order algorithm.  ...  In our experiments on a broadcast conversations recognition task, stochastic methods yield competitive results after only a short training period, but when spending enough computational resources for parallelization  ...  In general, optimization methods for machine learning can be subdivided into two categories: batch algorithms and stochastic algorithms.  ... 
doi:10.1109/icassp.2013.6639010 dblp:conf/icassp/WieslerRSN13 fatcat:yuv24pgmkbhwbhexbxmtjzv6ei

An Online Stochastic Kernel Machine for Robust Signal Classification [article]

Raghu G. Raj
2019 arXiv   pre-print
We present a novel variation of online kernel machines in which we exploit a consensus based optimization mechanism to guide the evolution of decision functions drawn from a reproducing kernel Hilbert  ...  SUMMARY In this paper we have introduced novel processing structures for incorporating a consensus based optimization mechanism into the evolution of decision functions drawn from a RKHS, derived a simple  ...  Whereas in the former setting, learning is based on a random batch of training samples from which a single hypothesis is formulated for prediction, in the online setting the learning algorithm observes  ... 
arXiv:1905.07686v2 fatcat:ctg7radssndhjhswpjnn2yutye

Partially Observed Markov Decision Processes: From Filtering to Controlled Sensing [Bookshelf]

Bo Wahlberg
2019 IEEE Control Systems  
Two widely used classes of reinforcement learning algorithms are Q-learning (a Robbins Munro stochastic approximation algorithm for estimating the value function) and the policy gradient algorithm (a stochastic  ...  Part 4 provides a concise description of stochastic optimization and reinforcement learning algorithms for POMDPs.  ... 
doi:10.1109/mcs.2019.2913493 fatcat:2eozixvhbjhzbby27rxrgqh4oa

First-Order Optimization (Training) Algorithms in Deep Learning

Oleg Rudenko, Oleksandr Bezsonov, Kyrylo Oliinyk
2020 International Conference on Computational Linguistics and Intelligent Systems  
Studies show that for this task a simple gradient descent algorithm is quite effective.  ...  In contrast to the problem of determining the structure, which is a discrete optimization (combinatorial), the search for optimal parameters is carried out in continuous space using some optimization methods  ...  However, the SAG technique can be utilized only with the smooth loss function and a convex objective function.  ... 
dblp:conf/colins/RudenkoBO20 fatcat:urkwrrkkq5fqvjcrxkmto4move

Stochastic Gradient Descent Algorithm for Glaucoma Detection using Frequency Domain Features of Retinal Images

2020 European Journal of Molecular and Clinical Medicine  
The input retinal images are given to frequency domain for feature extraction and SGD algorithm is used for detection. Experimental results show the performance of proposed system.  ...  Glaucoma is one of the major blindness causes for people aged 60 or over. The glaucoma detection using Stochastic Gradient Descent (SGD) algorithm is described in this study.  ...  Methods and Materials Stochastic gradient descent is a common and frequent algorithm used in numerous algorithms for machine learning. Primarily it forms the backbone of Neural Networks.  ... 
doi:10.31838/ejmcm.07.09.147 fatcat:d7zjia56hnf4ziriwr3w7gbbxa

Nonlinear Inverse Reinforcement Learning with Gaussian Processes

Sergey Levine, Zoran Popovic, Vladlen Koltun
2011 Neural Information Processing Systems  
We present a probabilistic algorithm for nonlinear inverse reinforcement learning.  ...  While most prior inverse reinforcement learning algorithms represent the reward as a linear combination of a set of features, we use Gaussian processes to learn the reward as a nonlinear function, while  ...  Ng and Krishnamurthy Dvijotham for helpful feedback and discussion. This work was supported by NSF Graduate Research Fellowship DGE-0645962.  ... 
dblp:conf/nips/LevinePK11 fatcat:b2zdqjxi4jhgvcnqou5wisoc2y

An Empirical Evaluation of Sequence-Tagging Trainers [article]

P. Balamurugan, Shirish Shevade, S. Sundararajan, S. S Keerthi
2013 arXiv   pre-print
On large datasets, online algorithms are preferred as batch learning methods are slow. These online algorithms were designed to solve either a primal or a dual problem.  ...  With this aim, we compare different algorithms and make recommendations, useful for a practitioner.  ...  Collins (2002) proposed the perceptron algorithm for structural learning.  ... 
arXiv:1311.2378v1 fatcat:v5l4au2qnzbplar4hzf5yeufby

A probabilistic fuzzy logic system: Learning in the stochastic environment with incomplete dynamics

Han-Xiong Li, Zhi Liu
2009 2009 IEEE International Conference on Systems, Man and Cybernetics  
Integrated into the fuzzy-PID structure, it will turn into a probabilistic fuzzy logic controller for the stochastic control.  ...  Using a unique threedimensional membership function (fuzz grade, time and probability), the probabilistic processing features can be added into the existing fuzzy configuration to construct a probabilistic  ...  It will be interesting to develop a more simple computational structure for probabilistic defuzzification.  ... 
doi:10.1109/icsmc.2009.5346199 dblp:conf/smc/LiL09 fatcat:wft3ljcxabaslglub3i3zblm5y

Page 1998 of Mathematical Reviews Vol. , Issue 84e [page]

1984 Mathematical Reviews  
Author’s summary: “Learning behaviors of variable-structure stochastic automata under a multiteacher environment are considered.  ...  As an extended form of the absolutely expedient learning algorithm, a general class of nonlinear learning algorithms, called the GAE scheme, is proposed as a reinforcement scheme in a multiteacher environment  ... 

Stochastic Learning Algorithms For Modeling Human Category Learning

Toshihiko Matsuka, James E. Corter
2007 Zenodo  
In this paper we introduce stochastic learning algorithms for NN models of human category learning and show that use of the algorithms can result in (a) rapid changes in accuracy and attention allocation  ...  Most neural network (NN) models of human category learning use a gradient-based learning method, which assumes that locally-optimal changes are made to model parameters on each learning trial.  ...  ACKNOWLEDGMENT Authors thank Stephen Jose Hanson, Catherine Hanson, Yasuaki Sakamoto, Areti Chouchourelou, and researchers at RUMBA for their helpful comments and suggestions.  ... 
doi:10.5281/zenodo.1054751 fatcat:ad6vfp2zgjetbkz7yhqsmg3cxm

Marginalized Kernels for RNA Sequence Data Analysis

Taishin Kin, Koji Tsuda, Kiyoshi Asai
2002 Genome Informatics Series  
The latter employs stochastic context-free grammar (SCFG) for estimating the secondary structure. We call the latter the marginalized count kernel (MCK).  ...  One is for RNA sequences with known secondary structures, the other for those without known secondary structures.  ...  Acknowledgment This work is partially supported by the Grant-in-Aid for Scientific Research on Priority Areas C "Genome Information Science ."  ... 
doi:10.11234/gi1990.13.112 fatcat:t6a3qtrqzbe2fjils56mxyndoi

Stochastic model for curvilinear structure reconstruction using morphological profiles

Seong-Gyun Jeong, Yuliya Tarabalka, Josiane Zerubia
2015 2015 IEEE International Conference on Image Processing (ICIP)  
In this work, we propose a stochastic model for curvilinear structure reconstruction using morphological profiles of path opening operator.  ...  Then, we formulate a stochastic optimization problem that detects line segments corresponding to the latent curvilinear structure in a scene.  ...  We proposed a new curvilinear structure reconstruction algorithm which exploits morphological profiles of path opening operator.  ... 
doi:10.1109/icip.2015.7351470 dblp:conf/icip/JeongTZ15 fatcat:qprp2v6i3veiro6jhgibscxylq

Segmentation of thalamus from MR images via task-driven dictionary learning

Luoluo Liu, Jeffrey Glaister, Xiaoxia Sun, Aaron Carass, Trac D. Tran, Jerry L. Prince, Martin A. Styner, Elsa D. Angelini
2016 Medical Imaging 2016: Image Processing  
In this work, we introduce a task-driven dictionary learning framework to find the optimal dictionary given a set of eleven features obtained from T1-weighted MRI and diffusion tensor imaging.  ...  In this dictionary learning framework, a linear classifier is designed concurrently to classify voxels as belonging to the thalamus or non-thalamus class.  ...  The dictionary is learned from a set of features extracted from T1-w and DTI MRI.  ... 
doi:10.1117/12.2214206 pmid:27601772 pmcid:PMC5010870 dblp:conf/miip/LiuGSCTP16 fatcat:dihcj7ggxndm5ckkqqthvvjupm
« Previous Showing results 1 — 15 out of 118,324 results