Feature selection using linear classifier weights

Mladenić, Dunja; Brank, Janez; Grobelnik, Marko; Milić-Frayling, Nataša

doi:10.1145/1008992.1009034

Cited by 158 publications

(94 citation statements)

References 6 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…A weight-based ranking was applied (Mladeni et al, 2004). Vector elements associated with small magnitude weights |Ŵ| are relatively unimportant for classification; where |Ŵ| the absolute value of is |Ŵ| normalized to the observed maximum.…”

Section: Feature Reductionmentioning

confidence: 99%

Alzheimer's disease diagnosis in individual subjects using structural MR images: Validation studies

et al. 2008

View full text Add to dashboard Cite

show abstract

Section: Feature Reductionmentioning

confidence: 99%

Alzheimer's disease diagnosis in individual subjects using structural MR images: Validation studies

et al. 2008

View full text Add to dashboard Cite

show abstract

“…Do's method uses Daubechies 4-tap wavelets and a pyramidal transform; the same weight is assigned to each subband and the low-frequency subband is discarded. The proposed method is also compared to a retrieval method based on Zernike moments (Khotanzad and Hong, 1990), in which the influence of each moment in the distance measure is weighted using an SVM (Mladenić et al, 2004). The results are reported on figure 12.…”

Section: Resultsmentioning

confidence: 99%

Wavelet optimization for content-based image retrieval in medical databases

Quellec

Lamard

Cazuguel

et al. 2010

Medical Image Analysis

175

View full text Add to dashboard Cite

We propose in this article a Content-Based Image Retrieval method for diagnosis aid in medical fields. In the proposed system, images are indexed in a generic fashion, without extracting domain-specific features: a signature is built for each image from its wavelet transform. These image signatures characterize the distribution of wavelet coefficients in each subband of the decomposition. A distance measure is then defined to compare two image signatures and thus retrieve the most similar images in a database when a query image is submitted by a physician. To retrieve relevant images from a medical database, the signatures and the distance measure must be related to the medical interpretation of images. As a consequence, we introduce several degrees of freedom in the system so that it can be tuned to any pathology and image modality. In particular, we propose to adapt the wavelet basis, within the lifting scheme framework, and to use a custom decomposition scheme. Weights are also introduced between subbands. All these parameters are tuned by an optimization procedure, using the medical grading of each image in the database to define a performance measure. The system is assessed on two medical image databases: one for diabetic retinopathy follow up and one for screening mammography, as well as a general purpose database. Results are promising: a mean precision of 56.50%, 70.91% and 96.10% is achieved for these three databases, when five images are returned by the system.

show abstract

“…Note that F is the set of all features taken into account. There are numerous feature weighting methods to assign weight to features like information gain [7], weights from a linear classifier [15], odds ratio, etc. Herein, we consider two well-known weighting schemas.…”

Section: B Feature Selection and Feature Weightingmentioning

confidence: 99%

Keyword Query Expansion on Linked Data Using Linguistic and Semantic Features

Shekarpour

Höffner

Lehmann

et al. 2013

2013 IEEE Seventh International Conference on Semantic Computing

View full text Add to dashboard Cite

Abstract-Effective search in structured information based on textual user input is of high importance in thousands of applications. Query expansion methods augment the original query of a user with alternative query elements with similar meaning to increase the chance of retrieving appropriate resources. In this work, we introduce a number of new query expansion features based on semantic and linguistic inferencing over Linked Open Data. We evaluate the effectiveness of each feature individually as well as their combinations employing several machine learning approaches. The evaluation is carried out on a training dataset extracted from the QALD question answering benchmark. Furthermore, we propose an optimized linear combination of linguistic and lightweight semantic features in order to predict the usefulness of each expansion candidate. Our experimental study shows a considerable improvement in precision and recall over baseline approaches.

show abstract

Feature selection using linear classifier weights

Cited by 158 publications

References 6 publications

Alzheimer's disease diagnosis in individual subjects using structural MR images: Validation studies

Alzheimer's disease diagnosis in individual subjects using structural MR images: Validation studies

Wavelet optimization for content-based image retrieval in medical databases

Keyword Query Expansion on Linked Data Using Linguistic and Semantic Features

Contact Info

Product

Resources

About