Numerical Analysis and Scientific Computing Commons^™

Direct Neighbor Search, Jilian Zhang, Kyriakos Mouratidis, Hwee Hwa Pang

Research Collection School Of Computing and Information Systems

In this paper we study a novel query type, called direct neighbor query. Two objects in a dataset are direct neighbors (DNs) if a window selection may exclusively retrieve these two objects. Given a source object, a DN search computes all of its direct neighbors in the dataset. The DNs define a new type of affinity that differs from existing formulations (e.g., nearest neighbors, nearest surrounders, reverse nearest neighbors, etc.) and finds application in domains where user interests are expressed in the form of windows, i.e., multi-attribute range selections. Drawing on key properties of the DN relationship, we develop an …

Semantic Visualization For Spherical Representation, Tuan M. V. Le, Hady W. Lauw

Research Collection School Of Computing and Information Systems

Visualization of high-dimensional data such as text documents is widely applicable. The traditional means is to find an appropriate embedding of the high-dimensional representation in a low-dimensional visualizable space. As topic modeling is a useful form of dimensionality reduction that preserves the semantics in documents, recent approaches aim for a visualization that is consistent with both the original word space, as well as the semantic topic space. In this paper, we address the semantic visualization problem. Given a corpus of documents, the objective is to simultaneously learn the topic distributions as well as the visualization coordinates of documents. We propose …

Diversity-Oriented Bi-Objective Hyper-Heuristics For Patrol Scheduling, Mustafa Misir, Hoong Chuin Lau

Research Collection School Of Computing and Information Systems

The patrol scheduling problem is concerned with assigning security teams to different stations for distinct time intervals while respecting a limited number of contractual constraints. The objective is to minimise the total distance travelled while maximising the coverage of the stations with respect to their security requirement levels. This paper introduces a hyper-heuristic strategy focusing on generating diverse solutions for a bi-objective patrol scheduling problem. While a variety of hyper-heuristics have been applied to a large suite of problem domains usually in the form of single-objective optimisation, we suggest an alternative approach for solving the patrol scheduling problem with two …

Interpretable Machine Learning And Sparse Coding For Computer Vision, Will Landecker

Dissertations and Theses

Machine learning offers many powerful tools for prediction. One of these tools, the binary classifier, is often considered a black box. Although its predictions may be accurate, we might never know why the classifier made a particular prediction. In the first half of this dissertation, I review the state of the art of interpretable methods (methods for explaining why); after noting where the existing methods fall short, I propose a new method for a particular type of black box called additive networks. I offer a proof of trustworthiness for this new method (meaning a proof that my method does not …

Online Multiple Kernel Regression, Doyen Sahoo, Steven C. H. Hoi, Bin Li

Research Collection School Of Computing and Information Systems

Kernel-based regression represents an important family of learning techniques for solving challenging regression tasks with non-linear patterns. Despite being studied extensively, most of the existing work suffers from two major drawbacks: (i) they are often designed for solving regression tasks in a batch learning setting, making them not only computationally inefficient and but also poorly scalable in real-world applications where data arrives sequentially; and (ii) they usually assume a fixed kernel function is given prior to the learning task, which could result in poor performance if the chosen kernel is inappropriate. To overcome these drawbacks, this paper presents a novel …

A Fast Decomposition Approach For Traffic Control, Xiaocheng Tang, Sébastien Blandin, Laura Wynter

Research Collection School Of Computing and Information Systems

Real-time road traffic control has been the subject of active research efforts for more than fifty years. In recent years, however, the convergence of ubiquitous sensing with seamless communication technologies has motivated the development of more computationally efficient control methods, able to operate in real-time in a live environment. In this work, we present a fast decomposition method for network optimization problems, with application to real-time traffic control. Our approach is based on a nonlinear programming formulation of the network control problem and consists of an alternating directions method using forward numerical simulation in place of one of the optimization …

Automated Prediction Of Glasgow Outcome Scale For Traumatic Brain Injury, Bolan Su, Thien Anh Dinh, A. K. Ambastha, Tianxia Gong, Tomi Silander, Shijian Lu, C. C. Tchoyoson Lim, Boon Chuan Pang, Cheng Kiang Lee, Tze-Yun Leong, Chew Lim Tan

Research Collection School Of Computing and Information Systems

Clinical features found in brain CT scan images are widely used in traumatic brain injury (TBI) as indicators for Glasgow Outcome Scale (GOS) prediction. However, due to the lack of automated methods to measure and quantify the CT scan image features, the computerized prediction of GOS in TBI has not been well studied. This paper introduces an automated GOS prediction system for traumatic brain CT images. Different from most existing systems that perform the prognosis based on pre-processed data, our system directly works on brain CT scan images based on the image features. Our system can also be extended to …

A Mathematical Model And Metaheuristics For Time Dependent Orienteering Problem, Aldy Gunawan, Zhi Yuan, Hoong Chuin Lau

Research Collection School Of Computing and Information Systems

This paper presents a generalization of the Orienteering Problem, the Time-Dependent Orienteering Problem (TDOP) which is based on the real-life application of providing automatic tour guidance to a large leisure facility such as a theme park. In this problem, the travel time between two nodes depends on the time when the trip starts. We formulate the problem as an integer linear programming (ILP) model. We then develop various heuristics in a step by step fashion: greedy construction, local search and variable neighborhood descent, and two versions of iterated local search. The proposed metaheuristics were tested on modified benchmark instances, randomly …

Evidence For Magnetic Clusters In Ni₁₋ₓvₓ Close To The Quantum Critical Concentration, R. Wang, S. Ubaid-Kassis, A. Schroeder, P. J. Baker, F. L. Pratt, S. J. Blundell, T. Lancaster, I. Franke, J. S. Möller, Thomas Vojta

Physics Faculty Research & Creative Works

The d-metal alloy Ni_1-xV_x undergoes a quantum phase transition from a ferromagnetic ground state to a paramagnetic ground state as the vanadium concentration x is increased. We present magnetization, ac-susceptibility and muon-spin relaxation data at several vanadium concentrations near the critical concentration x_c ~ 11.6 % at which the onset of ferromagnetic order is suppressed to zero temperature. Below x_c, the muon data reveal a broad magnetic field distribution indicative of a long-range ordered ferromagnetic state with spatial disorder. We show evidence of magnetic clusters in the ferromagnetic phase and close to the phase …

Rare Regions And Griffiths Singularities At A Clean Critical Point: The Five-Dimensional Disordered Contact Process, Thomas Vojta, John Igo, José A. Hoyos

Physics Faculty Research & Creative Works

We investigate the nonequilibrium phase transition of the disordered contact process in five space dimensions by means of optimal fluctuation theory and Monte Carlo simulations. We find that the critical behavior is of mean-field type, i.e., identical to that of the clean five-dimensional contact process. It is accompanied by off-critical power-law Griffiths singularities whose dynamical exponent z' saturates at a finite value as the transition is approached. These findings resolve the apparent contradiction between the Harris criterion, which implies that weak disorder is renormalization-group irrelevant, and the rare-region classification, which predicts unconventional behavior. We confirm and illustrate our theory by …

Structure Preserving Large Imagery Reconstruction, Ju Shen, Jianjun Yang, Sami Taha Abu Sneineh, Bryson Payne, Markus Hitz

Computer Science Faculty Publications

With the explosive growth of web-based cameras and mobile devices, billions of photographs are uploaded to the internet. We can trivially collect a huge number of photo streams for various goals, such as image clustering, 3D scene reconstruction, and other big data applications. However, such tasks are not easy due to the fact the retrieved photos can have large variations in their view perspectives, resolutions, lighting, noises, and distortions. Furthermore, with the occlusion of unexpected objects like people, vehicles, it is even more challenging to find feature correspondences and reconstruct realistic scenes. In this paper, we propose a structure-based image …

Theoretical And Experimental Investigation Of (E, 2e) Ionization Of Argon 3p In Asymmetric Kinematics At Intermediate Energy, Sadek Amami, Melike Ulu, Zehra Nur Ozer, Murat Yavuz, Suay Kazgoz, Mevlut Dogan, Oleg Zatsarinny, Klaus Bartschat, Don H. Madison

Physics Faculty Research & Creative Works

The field of electron-impact ionization of atoms, or (e, 2e), has provided significant detailed information about the physics of collisions. For ionization of hydrogen and helium, essentially exact numerical methods have been developed which can correctly predict what will happen. For larger atoms, we do not have theories of comparable accuracy. Considerable attention has been given to ionization of inert gases and, of the inert gases, argon seems to be the most difficult target for theory. There have been several studies comparing experiment and perturbative theoretical approaches over the last few decades, and generally qualitative but not quantitative agreement is …

Improving Fmri Analysis And Mr Reconstruction With The Incorporation Of Mr Relaxivities And Correlation Effect Examination, Meryem Muge Karaman

Dissertations (1934 -)

Functional magnetic resonance imaging (fMRI) and functional connectivity MRI (fcMRI) use the physical principles of nuclear MR to provide high resolution representations of brain activity and connectivity. As the fMRI and fcMRI signals are detected from the excited hydrogen atoms in a magnetic field, the acquired data is determined by the underlying physical processes, such as the MR relaxivities. In fMRI and fcMRI, the Fourier encoded frequency space measurements are reconstructed into brain images, then spatiotemporal processing operations are applied before computing the brain activation and connectivity statistics. This dissertation seeks to utilize the magnetic resonance (MR) relaxivities at different …

Understanding The Paradigm Shift To Computational Social Science In The Presence Of Big Data, Ray M. Chang, Robert J. Kauffman, Young Ok Kwon

Research Collection School Of Computing and Information Systems

The era of big data has created new opportunities for researchers to achieve high relevance and impact amid changes and transformations in how we study social science phenomena. With the emergence of new data collection technologies, advanced data mining and analytics support, there seems to be fundamental changes that are occurring with the research questions we can ask, and the research methods we can apply. The contexts include social networks and blogs, political discourse, corporate announcements, digital journalism, mobile telephony, home entertainment, online gaming, financial services, online shopping, social advertising, and social commerce. The changing costs of data collection and …

Manifold Learning For Jointly Modeling Topic And Visualization, Tuan Minh Van Le, Hady W. Lauw

Research Collection School Of Computing and Information Systems

Classical approaches to visualization directly reduce a document's high-dimensional representation into visualizable two or three dimensions, using techniques such as multidimensional scaling. More recent approaches consider an intermediate representation in topic space, between word space and visualization space, which preserves the semantics by topic modeling. We call the latter semantic visualization problem, as it seeks to jointly model topic and visualization. While previous approaches aim to preserve the global consistency, they do not consider the local consistency in terms of the intrinsic geometric structure of the document manifold. We therefore propose an unsupervised probabilistic model, called Semafore, which aims to …

Soml: Sparse Online Metric Learning With Application To Image Retrieval, Xingyu Gao, Steven C. H. Hoi, Yongdong Zhang, Ji Wan, Jintao Li

Research Collection School Of Computing and Information Systems

Image similarity search plays a key role in many multimedia applications, where multimedia data (such as images and videos) are usually represented in high-dimensional feature space. In this paper, we propose a novel Sparse Online Metric Learning (SOML) scheme for learning sparse distance functions from large-scale high-dimensional data and explore its application to image retrieval. In contrast to many existing distance metric learning algorithms that are often designed for low-dimensional data, the proposed algorithms are able to learn sparse distance metrics from high-dimensional data in an efficient and scalable manner. Our experimental results show that the proposed method achieves better …

Learning Relative Similarity By Stochastic Dual Coordinate Ascent, Pengcheng Wu, Ding Yi, Peilin Zhao, Chunyan Miao, Steven C. H. Hoi

Research Collection School Of Computing and Information Systems

Learning relative similarity from pairwise instances is an important problem in machine learning and has a wide range of applications. Despite being studied for years, some existing methods solved by Stochastic Gradient Descent (SGD) techniques generally suffer from slow convergence. In this paper, we investigate the application of Stochastic Dual Coordinate Ascent (SDCA) technique to tackle the optimization task of relative similarity learning by extending from vector to matrix parameters. Theoretically, we prove the optimal linear convergence rate for the proposed SDCA algorithm, beating the well-known sublinear convergence rate by the previous best metric learning algorithms. Empirically, we conduct extensive …

Predicting The Popularity Of Web 2.0 Items Based On User Comments, Xiangnan He, Ming Gao, Min-Yen Kan, Yiqun Liu, Kazunari Sugiyama

Research Collection School Of Computing and Information Systems

In the current Web 2.0 era, the popularity of Web resources fluctuates ephemerally, based on trends and social interest. As a result, content-based relevance signals are insufficient to meet users' constantly evolving information needs in searching for Web 2.0 items. Incorporating future popularity into ranking is one way to counter this. However, predicting popularity as a third party (as in the case of general search engines) is difficult in practice, due to their limited access to item view histories. To enable popularity prediction externally without excessive crawling, we propose an alternative solution by leveraging user comments, which are more accessible …

Changes On Cran, Kurt Hornik, Achim Zeileis

The R Journal

New packages in CRAN task views

New contributed packages

Other changes

The R Journal (June 2014) 6(1): Complete Issue, The R Foundation

The R Journal

Editorial, Deepayan Sarkar

Contributed Research Articles

Taming PITCHf/x Data with XML2R and pitchRx, Carson Sievert

A Multiscale Test of Spatial Stationarity for Textured Images in R, Matthew A. Nunes, Sarah L. Taylor, and Idris A. Eckley

Stratified Weibull Regression Model for Interval-Censored Data, Xiangdong Gu, David Shapiro, Michael D. Hughes, and Raji Balasubramanian

brainR: Interactive 3 and 4D Images of High Resolution Neuroimage Data, John Muschelli, Elizabeth Sweeney, and Ciprian Crainiceanu

The RWiener Package: an R Package Providing Distribution Functions for the Wiener Diffusion Model, Dominik Wabersich, and Joachim Vandekerckhove

PivotalR: A Package for Machine Learning on Big Data, Hai …