Filters








63,215 Hits in 3.6 sec

Comparing Bayesian Models of Annotation

Silviu Paun, Bob Carpenter, Jon Chamberlain, Dirk Hovy, Udo Kruschwitz, Massimo Poesio
2018 Transactions of the Association for Computational Linguistics  
Traditionally, majority voting was used for 1, and coefficients of agreement for 2 and 3. Lately, model-based analysis of corpus annotations have proven better at all three tasks.  ...  of noise in the form of random (spammy) annotators.  ...  Office of Naval Research.  ... 
doi:10.1162/tacl_a_00040 fatcat:ekkl7vqqanar7jsvpm5xtb6gcq

USING GRAPHICAL MODELS AND GENOMIC EXPRESSION DATA TO STATISTICALLY VALIDATE MODELS OF GENETIC REGULATORY NETWORKS

ALEXANDER J. HARTEMINK, DAVID K. GIFFORD, TOMMI S. JAAKKOLA, RICHARD A. YOUNG
2000 Biocomputing 2001  
The models that we use are based on Bayesian networks and their extensions.  ...  When we extend the graph semantics to permit annotated edges, we are able to score models describing relationships at a finer degree of specification.  ...  One limitation of comparing regulatory network models is that human effort is needed to formulate the models being compared.  ... 
doi:10.1142/9789814447362_0042 fatcat:ol7lor4ndrdhvh7t65cej7o4jq

Semi-supervised Learning for Automatic Image Annotation Based on Bayesian Framework

Dongping Tian
2014 International Journal of Control and Automation  
The novelty of our method mainly lies in two aspects: exploiting TSVM to improve the quality of training image dataset and utilizing the Bayesian model to predict the candidate annotations for the unseen  ...  On the other hand, a simple yet very efficient Bayesian model is built to implement image annotation by the maximum a posteriori (MAP) criterion.  ...  Bayesian Framework for Automatic Image Annotation For automatic image annotation, the Bayesian model works by finding the posterior probability that a concept belongs to an image.  ... 
doi:10.14257/ijca.2014.7.6.21 fatcat:2kwime5rkbc3vgr4nql5n6bapy

Leveraging Crowdsourcing Data for Deep Active Learning An Application

Jie Yang, Thomas Drake, Andreas Damianou, Yoelle Maarek
2018 Proceedings of the 2018 World Wide Web Conference on World Wide Web - WWW '18  
Experiments show that our framework can accurately learn annotator expertise, infer true labels, and effectively reduce the amount of annotations in model training as compared to state-of-the-art approaches  ...  Finally, our framework exploits the uncertainty of the deep learning model during prediction as well as the annotators' estimated expertise to minimize the number of required annotations and annotators  ...  the Bayesian DL model and the LFTC model, as it is a function of the DL discriminator and the model of annotators.  ... 
doi:10.1145/3178876.3186033 dblp:conf/www/YangDDM18 fatcat:6ngqcgzjnrc6tg47tmhbfcmwri

A Bayesian Evaluation Framework for Subjectively Annotated Visual Recognition Tasks [article]

Derek S. Prijatelj
2021 arXiv   pre-print
Machine learning-based predictors for these tasks rely on supervised training that models the behavior of the annotators, i.e., what would the average person's judgement be for an image?  ...  An interesting development in automatic visual recognition has been the emergence of tasks where it is not possible to assign objective labels to images, yet still feasible to collect annotations that  ...  who provided critiques on the drafts of this paper.  ... 
arXiv:2007.06711v2 fatcat:m4itv4nbmbhypjd4sr4y2zhzwa

Baum-Welch Style EM Approach on Simple Bayesian Models forWeb Data Annotation

Fatih Gelgi, Hasan Davulcu
2007 IEEE/WIC/ACM International Conference on Web Intelligence (WI'07)  
The extracted information has a certain level of accuracy which can be surpassed by using statistical models that are capable of contextual reasoning such as Bayesian models.  ...  Our contribution is an EM algorithm that operates on simple Bayesian models to re-annotate WAD.  ...  The final accuracies of EM models have been compared with simple Bayesian models without EM. The results are given in 3. NBC-EM and SBC-EM are the EM models of NBC and SBC.  ... 
doi:10.1109/wi.2007.12 dblp:conf/webi/GelgiD07 fatcat:pyy5e5p5ffdxpau5lp4ixim2tq

Baum-Welch Style EM Approach on Simple Bayesian Models forWeb Data Annotation

Fatih Gelgi, Hasan Davulcu
2007 IEEE/WIC/ACM International Conference on Web Intelligence (WI'07)  
The extracted information has a certain level of accuracy which can be surpassed by using statistical models that are capable of contextual reasoning such as Bayesian models.  ...  Our contribution is an EM algorithm that operates on simple Bayesian models to re-annotate WAD.  ...  The final accuracies of EM models have been compared with simple Bayesian models without EM. The results are given in 3. NBC-EM and SBC-EM are the EM models of NBC and SBC.  ... 
doi:10.1109/wi.2007.4427182 fatcat:6mqwj5t3qvecrgd6iykix63lgu

Bayesian Methods for Semi-supervised Text Annotation [article]

Kristian Miok, Gregor Pirs, Marko Robnik-Sikonja
2020 arXiv   pre-print
A recently proposed Bayesian ensemble method helps us to combine the annotators' labels with predictions of trained models.  ...  To alleviate the problem, we propose two semi-supervised methods to guide the annotation process: a Bayesian deep learning model and a Bayesian ensemble method.  ...  Improving Annotations using Bayesian Ensembles We propose a Bayesian ensemble as a support method for the annotation process.  ... 
arXiv:2010.14872v1 fatcat:e6qi7t5bpfhullphm4j6t6f4q4

Bayesian Loss for Crowd Count Estimation With Point Supervision

Zhiheng Ma, Xing Wei, Xiaopeng Hong, Yihong Gong
2019 2019 IEEE/CVF International Conference on Computer Vision (ICCV)  
On the contrary, we propose Bayesian loss, a novel loss function which constructs a density contribution probability model from the point annotations.  ...  Most of the state-of-the-art methods are based on density map estimation, which convert the sparse point annotations into a "ground truth" density map through a Gaussian kernel, and then use it as the  ...  On the contrary, we propose Bayesian loss, which constructs a density contribution probability model from the point annotations.  ... 
doi:10.1109/iccv.2019.00624 dblp:conf/iccv/MaWHG19 fatcat:hf2ztjgw3vdi3cuwsufbf4ca6u

Improving genome annotations using phylogenetic profile anomaly detection

T. S. Mikkelsen, J. E. Galagan, J. P. Mesirov
2004 Bioinformatics  
Results: We show that a probabilistic model of phylogenetic profiles, trained from a database of curated genome annotations, can be used to reliably detect errors in new annotations.  ...  We have sought to develop a method for improving new annotations that can automatically synthesize and use the information available in a database of other annotated genomes.  ...  Given the amount of genomic variation across the bacterial kingdom, it is not immediately evident that comparing an initial annotation to the MPE profile is better than simply comparing it to the closest  ... 
doi:10.1093/bioinformatics/bti027 pmid:15374867 fatcat:jdbfjhhu7fbc5aoooc5bgqsmdq

Integrating Semantic Annotations in Bayesian Causal Models

Hector G. Ceballos, Francisco J. Cantú Ortiz
2007 International Workshop on Description Logics  
Semantic Bayesian Causal Models We introduce Semantic Bayesian Causal Models(SBCM) which integrate a causal model with a semantic layer into an intelligent agent.  ...  Introduction Probabilistic reasoning has been powered by the formalization of causality theory through Bayesian causal models [1] .  ... 
dblp:conf/dlog/CeballosO07 fatcat:k236hi6btvefvncd6ztumdsyqa

GENE FUNCTION PREDICTION BY THE MULTI-LAYERED CLASSIFIER WITH MULTIFEATURES
English

GANGMAN YI, JAEHEE JUNG
2011 International Journal of Bioinformatics Research  
A comparative analysis of our suggested model and other gene functional annotation systems shows that our model outperforms than others especially in terms of a number of correctly predicted proteins.  ...  The base-classifier serves a base of meta-classifier with Bayesian network model and meta-classifier plays role of classifying the designated GO term from the root node.  ...  Barutcuoglu et al. [6] also used a Bayesian network for the purpose of developing a multi-label annotation method, overcoming the shortcoming of inconstancy between the child and parent annotations by  ... 
doi:10.9735/0975-3087.3.2.214-220 fatcat:ycuaxqej7jekndobcux2qvqosi

Personalized medicine for mucositis: Bayesian networks identify unique gene clusters which predict the response to gamma-d-glutamyl-l-tryptophan (SCV-07) for the attenuation of chemoradiation-induced oral mucositis

Gil Alterovitz, Cynthia Tuthill, Israel Rios, Katharina Modelska, Stephen Sonis
2011 Oral Oncology  
Predictive Bayesian networks were identified and functional cluster analyses were performed. A specific 10 gene cluster was a critical contributor to the predictability of the dataset.  ...  We identified 107 genes that discriminated SCV-07 responders from non-responders using four models and applied Akaike Information Criteria (AIC) and Bayes Factor (BF) analysis to evaluate predictive accuracy  ...  Bayes Factor is the odds of the marginal likelihood of one model compared to another model. AIC or Akaike Information Criterion is a measure of the goodness of fit of an estimated statistical model.  ... 
doi:10.1016/j.oraloncology.2011.07.006 pmid:21824803 fatcat:hqep5wzccng5rlhqdn27vxdsti

Bayesian Joint Topic Modelling for Weakly Supervised Object Localisation [article]

Zhiyuan Shi, Timothy M. Hospedales, Tao Xiang
2017 arXiv   pre-print
We propose a novel framework based on Bayesian joint topic modelling.  ...  can resolve ambiguity and lead to better learning and localisation. (2) The Bayesian formulation of the model enables easy integration of prior knowledge about object appearance to compensate for limited  ...  Initial localisation Table 1 reports the initial annotation accuracy of our model compared with state-of-the-art. Our model shows superior performance on all datasets.  ... 
arXiv:1705.03372v1 fatcat:6gskbbogw5ejjfqsjvz5brw5wy

Bayesian Joint Topic Modelling for Weakly Supervised Object Localisation

Zhiyuan Shi, Timothy M. Hospedales, Tao Xiang
2013 2013 IEEE International Conference on Computer Vision  
We propose a novel framework based on Bayesian joint topic modelling.  ...  can resolve ambiguity and lead to better learning and localisation. (2) The Bayesian formulation of the model enables easy integration of prior knowledge about object appearance to compensate for limited  ...  Initial localisation Table 1 reports the initial annotation accuracy of our model compared with state-of-the-art. Our model shows superior performance on all datasets.  ... 
doi:10.1109/iccv.2013.371 dblp:conf/iccv/ShiHX13 fatcat:r2kkj7j72jfvzlcvpqyj5u2quy
« Previous Showing results 1 — 15 out of 63,215 results