20,499 Hits in 4.2 sec

A Bayesian Approach for Scene Interpretation with Integrated Hierarchical Structure [chapter]

Martin Drauschke, Wolfgang Förstner
2011 Lecture Notes in Computer Science  
Then, we use the hierarchy graph of regions to construct a conditional Bayesian network, where the probabilities of class occurrences in the hierarchy are used to improve the classification results of  ...  We show that our framework is able to learn models for several objects, such that we can reliably detect instances of them in other images.  ...  We use the region hierarchy and the block of features F to construct a Bayesian network, as visualized in fig. 2 .  ... 
doi:10.1007/978-3-642-23123-0_1 fatcat:2wiks6k7gzhp5lwksl64w67wwu

Bayesian Stress Testing of Models in a Classification Hierarchy [article]

Bashar Awwad Shiekh Hasan, Kate Kelly
2020 arXiv   pre-print
In this work we propose a Bayesian framework to model the interaction amongst models in such a hierarchy.  ...  Building a machine learning solution in real-life applications often involves the decomposition of the problem into multiple models of various complexity.  ...  The hierarchy being represented as a DAG makes it easy to model the solution as a Bayesian belief network, which in turn allows the use of the full functionalities of the Bayesian framework -for example  ... 
arXiv:2005.12327v1 fatcat:dlvbl52ezfaktlcsldarksuc5q

Learning in Compositional Hierarchies: Inducing the Structure of Objects from Data

Joachim Utans
1993 Neural Information Processing Systems  
The resulting model can be interpreted as a Bayesian Belief Network and also is in many respects similar to the stochastic visual grammar described by Mjolsness.  ...  At each node in the hierarchy, a probability distribution governing its parameters must be learned. The connections between nodes reflects the structure of the object.  ...  At OGI supported was provided in part under grant ONR N00014-92-J-4062. Discussions with S. Knerr, E. Mjolsness and S. Omohundro were helpful in preparing this work.  ... 
dblp:conf/nips/Utans93 fatcat:h7afvi3yarcepezfos7joce6qe

ACO-Based Bayesian Network Ensembles for the Hierarchical Classification of Ageing-Related Proteins [chapter]

Khalid M. Salama, Alex A. Freitas
2013 Lecture Notes in Computer Science  
In this paper we tackle the hierarchical classification problem in a local fashion, by learning an ensemble of Bayesian network classifiers for each class in the hierarchy and combining their outputs with  ...  The ensemble is built using ABC-Miner, our recently introduced Ant-based Bayesian Classification algorithm. We use different types of protein representations to learn different classification models.  ...  Carlos Silla for extracting the feature sets used in our experiments and Dr. Joao Pedro de Magalhaes for his valuable advice about the creation of the ageing-related protein's dataset.  ... 
doi:10.1007/978-3-642-37189-9_8 fatcat:jsek2syyizej5dgothhq7qkd7m

Constructing Hierarchical Bayesian Networks With Pooling

Kaneharu Nishino, Mary Inaba
Inspired by the Bayesian brain hypothesis and deep learning, we develop a Bayesian autoencoder, a method of constructing recognition systems using a Bayesian network.  ...  We construct hierarchical Bayesian networks based on feature extraction and implement pooling to achieve invariance within a Bayesian network framework.  ...  Acknowledgments This work was supported by JSPS Grant-in-Aid for JSPS Fellows, JP17J09110.  ... 
doi:10.1609/aaai.v32i1.12191 fatcat:jijul4li7ja6pnxqvoahlb6ll4

A fuzzy analytic hierarchy attribute weighting and deep learning for improving CHD prediction of optimized semi parametric extended dynamic bayesian network

K. Gomathi, D. Shanmuga Priyaa
2017 International Journal of Engineering & Technology  
In this paper, Optimized Semi parametric Extended Deep Dynamic Bayesian Network (OSEDDBN) is proposed by integrating deep learning architecture with OSEDBN to improve the ability of extracting more important  ...  The deep learning network is generated the various time points in the next level to improve the analysis and prediction of CHD.  ...  In this paper Deep Dynamic Bayesian Network (DDBN) and Fuzzy Analytic Hierarchy Process (Fuzzy-AHP) presented to improve prognosis of Coronary Heart Disease (CHD).  ... 
doi:10.14419/ijet.v7i1.1.9274 fatcat:2rl3cv4etrf2njaczjrk2fo3ea

Bayesian Aggregation For Hierarchical Genre Classification

Christopher DeCoro, Zafer Barutçuoglu, Rebecca Fiebrink
2007 Zenodo  
ACKNOWLEDGMENTS We would like to thank Cory McKay for sharing the Bodhidharma MIDI dataset used for evaluation of this work, and for his helpful advice and opinions.  ...  We use base classifications from the training set in order to generate a Bayesian network that describes the hierarchy.  ...  The class hierarchy (a) is transformed into a Bayesian network (b).  ... 
doi:10.5281/zenodo.1416012 fatcat:r2lzzisi5ja2zcv5t33zzpxpi4

Learning invariant features using inertial priors

Thomas Dean
2007 Annals of Mathematics and Artificial Intelligence  
Each component network has an associated receptive field corresponding to components residing in the level directly below it in the hierarchy.  ...  The resulting model is a hierarchical Bayesian network factored into modular component networks embedding variable-order Markov models.  ...  The resulting computational model takes the form of a Bayesian network [53, 35, 34] which is constructed from many smaller networks each capable of learning an invariant feature.  ... 
doi:10.1007/s10472-006-9039-9 fatcat:62dchpkavzgwdemsgwncpcap6y

Medical image modality classification using discrete Bayesian networks

Jacinto Arias, Jesus Martínez-Gómez, Jose A. Gámez, Alba G. Seco de Herrera, Henning Müller
2016 Computer Vision and Image Understanding  
classifier used in image classification.  ...  In this paper we propose a complete pipeline for medical image modality classification focused on the application of discrete Bayesian network classifiers.  ...  known as Bayesian Network Classifiers (BNCs) [39, 8] .  ... 
doi:10.1016/j.cviu.2016.04.002 fatcat:5z7hdsimfbf5nmfkrdgkp6vbmi

Nonparametric Variational Auto-encoders for Hierarchical Representation Learning [article]

Prasoon Goyal, Zhiting Hu, Xiaodan Liang, Chenyu Wang, Eric Xing
2017 arXiv   pre-print
Both the neural parameters and Bayesian priors are learned jointly using tailored variational inference.  ...  We apply our model in video representation learning.  ...  Then, each frame is passed through a pre-trained convolutional neural network (CNN) to obtain frame features. The resulting frame features are then used as sequence elements x mn in our framework.  ... 
arXiv:1703.07027v2 fatcat:fllrzupobrfhjpjeejjgzgmve4

Positive Feature Values Prioritized Hierarchical Redundancy Eliminated Tree Augmented Naive Bayes Classifier for Hierarchical Feature Spaces [article]

Cen Wan
2022 arXiv   pre-print
The Hierarchical Redundancy Eliminated Tree Augmented Naive Bayes (HRE-TAN) classifier is a semi-naive Bayesian model that learns a type of hierarchical redundancy-free tree-like feature representation  ...  In this work, we propose two new types of positive feature values prioritized hierarchical redundancy eliminated tree augmented naive Bayes classifiers that focus on features bearing positive instance  ...  In this wo vised learning tasks where hierarchic are used as features.  ... 
arXiv:2204.05668v1 fatcat:y5mse7e4lfephcf5ohfom46oqm

View Learning for Statistical Relational Learning: With an Application to Mammography

Jesse Davis, Elizabeth S. Burnside, Inês de Castro Dutra, David Page, Raghu Ramakrishnan, Vítor Santos Costa, Jude W. Shavlik
2005 International Joint Conference on Artificial Intelligence  
Such new fields or tables can also be highly useful in learning. We provide SRL with the capability of learning new views.  ...  A key capability of SRL is the learning of arcs (in the Bayes net sense) connecting entries in different rows of a relational table, or in different tables.  ...  Elizabeth Burnside is supported by a General Electric Research in Radiology Academic Fellowship. Inês Dutra and Vítor Santos Costa are on leave from Federal University of Rio de Janeiro, Brazil.  ... 
dblp:conf/ijcai/DavisBDPRCS05 fatcat:sz5uspzjerb2fgjhpfxrioawna

Deep Learning Face Representation from Predicting 10,000 Classes

Yi Sun, Xiaogang Wang, Xiaoou Tang
2014 2014 IEEE Conference on Computer Vision and Pattern Recognition  
When learned as classifiers to recognize about 10, 000 face identities in the training set and configured to keep reducing the neuron numbers along the feature extraction hierarchy, these deep ConvNets  ...  DeepID features are taken from the last hidden layer neuron activations of deep convolutional networks (ConvNets).  ...  We randomly choose 80% (4349) people from Celeb-Faces to learn the DeepID, and use the remaining 20% people to learn the face verification model (Joint Bayesian or neural networks).  ... 
doi:10.1109/cvpr.2014.244 dblp:conf/cvpr/SunWT14 fatcat:gdyfboylxjhs5of4dag6ckwznq

Robust Character Recognition Using a Hierarchical Bayesian Network [chapter]

John Thornton, Torbjorn Gustafsson, Michael Blumenstein, Trevor Hine
2006 Lecture Notes in Computer Science  
In the current study we extend Hawkins' work by comparing the performance of a backpropagation neural network with our own implementation of a hierarchical Bayesian network in the well-studied domain of  ...  The first partial proof of concept of Hawkins' model was recently developed using a hierarchically organised Bayesian network that was tested on a simple pattern recognition problem [3] .  ...  1 The Hierarchical Bayesian Network Figure 1 presents a simplified example of the hierarchical Bayesian network used in the current study.  ... 
doi:10.1007/11941439_157 fatcat:5q4fb4oldbdyhjwkvsjn3ekz2e

Embedding Visual Hierarchy With Deep Networks for Large-Scale Visual Recognition

Tianyi Zhao, Baopeng Zhang, Ming He, Wei Zhang, Ning Zhou, Jun Yu, Jianping Fan
2018 IEEE Transactions on Image Processing  
of object classes), and a Bayesian approach is used to adapt a pre-trained visual hierarchy automatically to the improvements of deep features (that are used for image and object class representation)  ...  In this paper, a level-wise mixture model (LMM) is developed by embedding visual hierarchy with deep networks to support large-scale visual recognition (i.e., recognizing thousands or even tens of thousands  ...  By integrating deep learning with multi-task learning, deep multi-task learning have received many attentions recently by using the deep networks to learn more representative features and integrating multi-task  ... 
doi:10.1109/tip.2018.2845118 pmid:29994211 fatcat:jzsf7otuirfg7alqvbtdy553pu
« Previous Showing results 1 — 15 out of 20,499 results