Filters








1,106,020 Hits in 6.1 sec

Relation Based Term Weighting Regularization [chapter]

Hao Wu, Hui Fang
2012 Lecture Notes in Computer Science  
In this work, we study how to utilize semantic relations among query terms to regularize term weighting.  ...  Traditional retrieval models compute term weights based on only the information related to individual terms such as TF and IDF. However, query terms are related.  ...  This material is based upon work supported by the National Science Foundation under Grant Number IIS-1017026 and HP Labs Innovation Research Program.  ... 
doi:10.1007/978-3-642-28997-2_10 fatcat:slihh2t6xnhm3fzynt26lpnqmu

Query Aspect Based Term Weighting Regularization in Information Retrieval [chapter]

Wei Zheng, Hui Fang
2010 Lecture Notes in Computer Science  
Despite its importance, the use of semantic relations among query terms for term weighting regularization has been under-explored in information retrieval.  ...  In this paper, we study the incorporation of query term relations into existing retrieval models and focus on addressing the challenge, i.e., how to regularize the weights of terms in different query aspects  ...  It remains unclear how to regularize term weighting based on query term relations and how to systematically incorporate the term weighting regularization functions into existing retrieval functions.  ... 
doi:10.1007/978-3-642-12275-0_31 fatcat:rwmivjantjf6vjilcboqxly5m4

Regular expression-based learning to extract bodyweight values from clinical notes

Maureen A. Murtaugh, Bryan Smith Gibson, Doug Redd, Qing Zeng-Treitler
2015 Journal of Biomedical Informatics  
This annotation served to develop the annotation process and identify terms associated with bodyweight related measures for training the supervised learning algorithm.  ...  In the dataset of weights from 3561 notes, 7.7% of notes contained bodyweight related measures that were not available as structured data.  ...  Regular expression based learning Regular expression-based learning has been an active area of research in computer science and to a lesser degree in biomedical informatics.  ... 
doi:10.1016/j.jbi.2015.02.009 pmid:25746391 fatcat:tf5nyxpaq5hbnmqqpwd3xogmwa

A Study of Concept-based Weighting Regularization for Medical Records Search

Yue Wang, Xitong Liu, Hui Fang
2014 Proceedings of the 52nd Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers)  
In particular, we apply axiomatic approaches and propose two weighting regularization methods that adjust the weighting based on the relations among the concepts.  ...  One commonly used approach is to apply NLP tools to map terms from queries and documents to concepts and then compute the relevance scores based on the conceptbased representation.  ...  In particular, we proposed two weighting regularization methods based on the relations among concepts.  ... 
doi:10.3115/v1/p14-1057 dblp:conf/acl/WangLF14 fatcat:hrvrlmleb5fcdfrxgr3jg55rvu

Recommender Systems with Characterized Social Regularization

Tzu-Heng Lin, Chen Gao, Yong Li
2018 Proceedings of the 27th ACM International Conference on Information and Knowledge Management - CIKM '18  
In this paper, we present a novel CSR (short for Characterized Social Regularization) model by designing a universal regularization term for modeling variable social influence.  ...  Existing social recommendation methods are based on the fact that users preference or decision is influenced by their social friends' behaviors.  ...  Social regularization is a regularization term considering social relation S. We denote the added term in social regularization as Social(Ω, S), where Ω denotes parameters of latent model.  ... 
doi:10.1145/3269206.3269234 dblp:conf/cikm/LinGL18 fatcat:ghzs3b34ibbydjp424e43ojblu

A Novel Multi-relation Regularization Method for Regression and Classification in AD Diagnosis [chapter]

Xiaofeng Zhu, Heung-Il Suk, Dinggang Shen
2014 Lecture Notes in Computer Science  
After feature selection based on the optimal weight coefficients, we train two support vector regression models to predict the clinical scores of Alzheimer's Disease Assessment Scale-Cognitive subscale  ...  In this paper, we consider the joint regression and classification in Alzheimer's disease diagnosis and propose a novel multi-relation regularization method that exploits the relational information inherent  ...  To this end, we devise a new regularization term with the claim that, if some features are related to each other, the same or the similar relation is expected to be preserved between the respective weight  ... 
doi:10.1007/978-3-319-10443-0_51 pmid:25320825 pmcid:PMC6892168 fatcat:go3sixpg7ngpnlalwizn4eqame

Accounting for Language Changes Over Time in Document Similarity Search

Sara Morsy, George Karypis
2016 ACM Transactions on Information Systems  
Given a query document, ranking the documents in a collection based on how similar they are to the query is an essential task with extensive applications.  ...  For example, many terms add or lose one or more senses to meet people's evolving needs.  ...  term-time-specific transition regularization weights is better than having the same weights in link regularized dynamic topic models.  ... 
doi:10.1145/2934671 fatcat:73eltun2fnettmzsbso3mnhngu

Gram Regularization for Multi-view 3D Shape Retrieval [article]

Zhaoqun Li
2020 arXiv   pre-print
To make up the gap, in this paper, we propose a novel regularization term called Gram regularization which reinforces the learning ability of the network by encouraging the weight kernels to extract different  ...  By forcing the variance between weight kernels to be large, the regularizer can help to extract discriminative features.  ...  Backpropagation of Gram regularization The Gram regularization is data independent and is calculated by weight group, so the gradient of each weight is only related to its corresponding weight group.  ... 
arXiv:2011.07733v1 fatcat:bay4tmaoqrhznkeb4n3yy7pfya

Semi-supervised learning with regularized Laplacian

K. Avrachenkov, P. Chebotarev, A. Mishenin
2016 Optimization Methods and Software  
We study a semi-supervised learning method based on the similarity graph and Regularized Laplacian.  ...  We demonstrate on numerical examples that the Regularized Laplacian method is robust with respect to the choice of the regularization parameter and outperforms the Laplacian-based heat kernel methods.  ...  In particular, we can conclude that in terms of robustness the Regularized Laplacian method is comparable in performance with the PageRank method and outperforms the related heat kernel based methods.  ... 
doi:10.1080/10556788.2016.1193176 fatcat:eai3cln2cvbxjdjsidjypxvxdm

Semi-supervised Learning with Regularized Laplacian [article]

Konstantin Avrachenkov, Pavel Chebotarev, Alexey Mishenin
2015 arXiv   pre-print
We study a semi-supervised learning method based on the similarity graph and RegularizedLaplacian.  ...  We give convenient optimization formulation of the Regularized Laplacian method and establishits various properties.  ...  In particular, we can conclude that the Regularized Laplacian method is comparable in performance with the PageRank based method and outperforms the related heat kernel based methods in terms of robustness  ... 
arXiv:1508.04906v1 fatcat:dffoyv2frfha7oeddes4c557vu

Graph Based Classification Methods Using Inaccurate External Classifier Information [article]

Sundararajan Sellamanickam, Sathiya Keerthi Selvaraj
2012 arXiv   pre-print
We also propose an efficient least squares regularization (LSR) based method and relate it to information regularization methods.  ...  We extend the LGC and weighted vote relational neighbor classification (WvRN) methods to support usage of external classifier information.  ...  The objective function consists of two terms. The first term known as the graph regularization (GR) term makes use of the weight matrix W and imposes smoothness of the function over the graph.  ... 
arXiv:1206.5915v1 fatcat:a2jkcc7ndfhwtnak2vonrr5p2a

Multilevel Models Allow Modular Specification of What and Where to Regularize, Especially in Small Area Estimation [article]

Michael Tzen
2018 arXiv   pre-print
The key idea is to use the design-based variance to control the amount of 'between' regularization and prior information to regularize the components 'within' a predictor.  ...  The goal is to let the design-based estimate have authority (when precise) but defer to a model-based prediction when imprecise.  ...  For this reason, the rest of this article will primarily use the term regularize / regularization instead of shrink / shrinkage.  ... 
arXiv:1805.08233v1 fatcat:k5m7mou7vfdalkvjoarbacleja

Automatic Spatially-Adaptive Balancing of Energy Terms for Image Segmentation [article]

Josna Rao, Ghassan Hamarneh, Rafeef Abugharbieh
2009 arXiv   pre-print
The objective function typically involves weights for balancing competing image fidelity and segmentation regularization cost terms.  ...  We propose a novel technique that autonomously balances these terms in a spatially-adaptive manner through the incorporation of image reliability in a graph-based segmentation framework.  ...  We begin by formulating the energy of a contour and specifying how the regularization term is weighted in our definition.  ... 
arXiv:0906.4131v2 fatcat:ovtr74q7gngkxbluhx6i6x4enm

Nomenclature and Terminology for Dendrimers with Regular Dendrons and for Hyperbranched Polymers

2017 Chemistry International  
With respect to dynamic development, particularly in the analysis and investigation of biomacromolecules, terms related to bioanalytical samples, enzymatic methods, immunoanalytical methods, methods used  ...  Nomenclature and Terminology for Dendrimers with Regular Dendrons and for Hyperbranched Polymers This document provides recommendations for (i) definitions of terms related to dendrimers with regular dendrons  ...  (CGPM) noted the intention of the International Committee of Weights and Measures (CIPM) to revise the entire International System of Units (SI) by linking all seven base units to seven fundamental physical  ... 
doi:10.1515/ci-2017-0224 fatcat:m2kgnqo5sbgo5l2jsncf45pyii

Adaptive Regularization via Residual Smoothing in Deep Learning Optimization [article]

Junghee Cho, Junseok Kwon, Byung-Woo Hong
2019 arXiv   pre-print
The degree of regularization at each element in the target space of the neural network architecture is determined based on the residual at each optimization iteration in an adaptive way.  ...  Our data-driven regularity is imposed by adaptively smoothing a simplified objective function in which the explicit regularization term is omitted in an alternating manner between the evaluation of residual  ...  Explicit Regularization Weight Decay: The objective function is assumed to include a regularization term that penalizes a perturbation of unknown parameters in terms of L 2 2 norm.  ... 
arXiv:1907.09750v2 fatcat:2futs44xvnbzhgxlwbywpwkb3m
« Previous Showing results 1 — 15 out of 1,106,020 results