Sejong Yoon

Mutual information-based SVM-RFE for diagnostic classification of digitized mammograms

Computer aided diagnosis (CADx) systems for digitized mammograms solve the problem of classification between benign and malignant tissues while studies have shown that using only a subset of features generated from the mammograms can yield higher classification accuracy. To this end, we propose a mutual information-based Support Vector Machine Recursive Feature Elimination (SVM-RFE) as the classification method with feature selection in this paper. We have conducted extensive experiments on publicly available mammographic data and the obtained results indicate that the proposed method outperforms other SVM and SVM-RFE-based methods.

AdaBoost-based multiple SVM-RFE for classification of mammograms in DDSM

BackgroundDigital mammography is one of the most promising options to diagnose breast cancer which is the most common cancer in women. However, its effectiveness is enfeebled due to the difficulty in distinguishing actual cancer lesions from benign abnormalities, which results in unnecessary biopsy referrals. To overcome this issue, computer aided diagnosis (CADx) using machine learning techniques have been studied worldwide. Since this is a classification problem and the number of features obtainable from a mammogram image is infinite, a feature selection method that is tailored for use in the CADx systems is needed.MethodsWe propose a feature selection method based on multiple support vector machine recursive feature elimination (MSVM-RFE). We compared our method with four previously proposed feature selection methods which use support vector machine as the base classifier. Experiments were performed on lesions extracted from the Digital Database of Screening Mammography, the largest public digital mammography database available. We measured average accuracy over 5-fold cross validation on the 8 datasets we extracted.ResultsSelecting from 8 features, conventional algorithms like SVM-RFE and multiple SVM-RFE showed slightly better performance than others. However, when selecting from 22 features, our proposed modified multiple SVM-RFE using boosting outperformed or was at least competitive to all others.ConclusionOur modified method may be a possible alternative to SVM-RFE or the original MSVM-RFE in many cases of interest. In the future, we need a specific method to effectively combine models trained during the feature selection process and a way to combine feature subsets generated from individual SVM-RFE instances.

Fast ADMM algorithm for distributed optimization with adaptive penalty

We propose new methods to speed up convergence of the Alternating Direction Method of Multipliers (ADMM), a common optimization tool in the context of large scale and distributed learning. The proposed method accelerates the speed of convergence by automatically deciding the constraint penalty needed for parameter consensus in each iteration. In addition, we also propose an extension of the method that adaptively determines the maximum number of iterations to update the penalty. We show that this approach effectively leads to an adaptive, dynamic network topology underlying the distributed optimization. The utility of the new penalty update schemes is demonstrated on both synthetic and real data, including a computer vision application of distributed structure from motion.

Relative spatial features for image memorability

Recent studies in image memorability showed that the memorability of an image is a measurable quantity and is closely correlated with semantic attributes. However, the intrinsic characteristics of memorability are not yet fully understood. It has been reported that in contrast to a popular belief unusualness or aesthetic beauty of the image may not be positively correlated with the image memorability. This counter-intuitive characteristic of memorability hinders a better understanding of image memorability and its applicability. In this paper, we investigate two new spatial features that are closely correlated with the image memorability yet intuitively explainable. We propose the Weighted Object Area (WOA) that jointly considers the location and size of objects and the Relative Area Rank (RAR) that captures the relative unusualness of the size of objects. We empirically demonstrate their useful correlation with the image memorability. Results show that both WOA and RAR can improve the memorability prediction. In addition, we provide evidence that the RAR can effectively capture object-centric unusualness of size.

Sentiment Flow for Video Interestingness Prediction

Computational analysis and prediction of digital media interestingness is a challenging task, largely driven by subjective nature of interestingness. Several attempts were made to construct a reliable measure and obtain a better understanding of interestingness based on various psychological study results. However, most current works focus on interestingness prediction for images. While the video affective analysis has been studied for quite some time, there are few works that explictly try to predict interestingness of videos. In this work, we extend a recent pilot study on the video interestingness prediction by using a mid-level representation of sentiment (emotion) sequence. We evaluate our proposed framework on three datasets including the datasets proposed by the pilot study and show that the result effectively verifies a promising utility of the approach.

k-Top Scoring Pair Algorithm for feature selection in SVM with applications to microarray data classification

Top Scoring Pair (TSP) and its ensemble counterpart, k-Top Scoring Pair (k-TSP), were recently introduced as competitive options for solving classification problems of microarray data. However, support vector machine (SVM) which was compared with these approaches is not equipped with feature or variable selection mechanism while TSP itself is a kind of variable selection algorithm. Moreover, an ensemble of SVMs should also be considered as a possible competitor to k-TSP. In this work, we conducted a fair comparison between TSP and SVM-recursive feature elimination (SVM-RFE) as the feature selection method for SVM. We also compared k-TSP with two ensemble methods using SVM as their base classifier. Results on ten public domain microarray data indicated that TSP family classifiers serve as good feature selection schemes which may be combined effectively with other classification methods.

Mass Lesions Classification in Digital Mammography using Optimal Subset of BI-RADS and Gray Level Features

Computer-aided diagnosis of mass lesions in Digital Database for Screening Mammography (DDSM) is investigated using a recently developed SVM based on recursive feature elimination (SVM-RFE) as the classification technique. To evaluate the generalizability, computer-aided diagnosis using cross-institutional mammograms is also examined. The results in this paper indicate that using only a subset of the available set of features facilitates increased computer-aided diagnosis accuracy, and that computer-aided diagnosis accuracy using cross-institutional mammograms is generally lower than when using same-institutional mammograms.

Decentralized approximate Bayesian inference for distributed sensor network

Bayesian models provide a framework for probabilistic modelling of complex datasets. Many such models are computationally demanding, especially in the presence of large datasets. In sensor network applications, statistical (Bayesian) parameter estimation usually relies on decentralized algorithms, in which both data and computation are distributed across the nodes of the network. In this paper we propose a framework for decentralized Bayesian learning using Bregman Alternating Direction Method of Multipliers (BADMM). We demonstrate the utility of our framework, with Mean Field Variational Bayes (MFVB) as the primitive for distributed affine structure from motion (SfM).

Filling in the blanks: reconstructing microscopic crowd motion from multiple disparate noisy sensors

Tracking the movement of individuals in a crowd is an indispensable component to reconstructing crowd movement, with applications in crowd surveillance and data-driven animation. Typically, multiple sensors are distributed over wide area and often they have incomplete coverage of the area or the input introduces noise due to the tracking algorithm or hardware failure. In this paper, we propose a novel refinement method that complements existing crowd tracking solutions to reconstruct a holistic view of the microscopic movement of individuals in a crowd, from noisy tracked data with missing and even incomplete information. Central to our approach is a global optimization based trajectory estimation with modular objective functions. We empirically demonstrate the potential utility of our approach in various scenarios that are standard in crowd dynamic analysis and simulations.

Multiple SVM-RFE Using Boosting for Mammogram Classification

Digital mammography is an effective method to diagnose breast cancer. However, unnecessary biopsies caused by low accuracy in classifying benign abnormalities and malignant ones are challenging problem of the approach. To resolve the issue, computer aided diagnosis (CADx) using various AI techniques have been proposed. Recently, reports indicate that CADx systems can be improved by exploiting mammogram and AI algorithm-specific feature selection schemes. In this regard, we propose a modified feature selection method based on a recently developed multiple support vector machine recursive feature elimination (MSVM-RFE). Experimental results on real world digital mammograms show that our method demonstrated competitive performances.