Numerical Analysis and Scientific Computing Commons^™

Rfit: Rank-Based Estimation For Linear Models, John D. Kloke, Joseph W. Mckeen

The R Journal

In the nineteen seventies, Jureĉková and Jaeckel proposed rank estimation for linear models. Since that time, several authors have developed inference and diagnostic methods for these estimators. These rank-based estimators and their associated inference are highly efficient and are robust to outliers in response space. The methods include estimation of standard errors, tests of general linear hypotheses, confidence intervals, diagnostic procedures including studentized residuals, and measures of influential cases. We have developed an R package, Rfit, for computing of these robust procedures. In this paper we highlight the main features of the pack age. The package uses standard linear …

It's Not What You Draw, It's What You Don't Draw, Paul Murrell

The R Journal

The R graphics engine has new support for drawing complex paths via the functions polypath() and grid.path(). This article explains what is meant by a complex path and demonstrates the usefulness of complex paths in drawing non-trivial shapes, logos, customised data symbols, and maps.

Debugging Grid Graphics, Paul Murrell, Velvet Ly

The R Journal

A graphical scene that has been produced using the grid graphics package consists of grobs (graphical objects) and viewports. This article describes functions that allow the exploration and inspection of the grobs and viewports in a grid scene, including several functions that are available in a new package called gridDe bug. The ability to explore the grobs and view ports in a grid scene is useful for adding more drawing to a scene that was produced using grid and for understanding and debugging the grid code that produced a scene.

The State Of Naming Conventions In R, Rasmus Bååth

The R Journal

Most programming language communities have naming conventions that are generally agreed upon, that is, a set of rules that governs how functions and variables are named. This is not the case with R, and a review of unofficial style guides and naming convention us age on CRAN shows that a number of different naming conventions are currently in use. Some naming conventions are, however, more popular than others and as a newcomer to the R community or as a developer of a new package this could be useful to consider when choosing what naming convention to adopt.

Visualization For Anomaly Detection And Data Management By Leveraging Network, Sensor And Gis Techniques, Zhaoxia Wang, Chee Seng Chong, Rick S. M. Goh, Wanqing Zhou, Dan Peng, Hoong Chor Chin

Research Collection School Of Computing and Information Systems

This paper studies the importance of visualization for discerning and interpreting patterns of data and its application for solving real problems, such as anomaly detection and data management. There are various ways to realize visualization to cater to the needs of numerous real life applications. Depending on needs, a combination of some of these ways may be required for presenting an effective visualization. The authors present visualization schemes for anomaly detection/condition monitoring and data management by leveraging network techniques and combining them with modern techniques such as sensor, database, mobile communication, GPS and GIS techniques. Two case studies are presented …

Design And Optimization Of Four-Dimensional Cone-Beam Computed To Mography In Image-Guided Radiation Therapy, Moiz Ahmad

Dissertations and Theses (Open Access)

The influence of respiratory motion on patient anatomy poses a challenge to accurate radiation therapy, especially in lung cancer treatment. Modern radiation therapy planning uses models of tumor respiratory motion to account for target motion in targeting. The tumor motion model can be verified on a per-treatment session basis with four-dimensional cone-beam computed tomography (4D-CBCT), which acquires an image set of the dynamic target throughout the respiratory cycle during the therapy session. 4D-CBCT is undersampled if the scan time is too short. However, short scan time is desirable in clinical practice to reduce patient setup time. This dissertation presents the …

A Survey Of Recommender Systems In Twitter, Su Mon Kywe, Ee Peng Lim, Feida Zhu

Research Collection School Of Computing and Information Systems

Twitter is a social information network where short messages or tweets are shared among a large number of users through a very simple messaging mechanism. With a population of more than 100M users generating more than 300M tweets each day, Twitter users can be easily overwhelmed by the massive amount of information available and the huge number of people they can interact with. To overcome the above information overload problem, recommender systems can be introduced to help users make the appropriate selection. Researchers have began to study recommendation problems in Twitter but their works usually address individual recommendation tasks. There …

Detecting Anomalies In Bipartite Graphs With Mutual Dependency Principles, Hanbo Dai, Feida Zhu, Ee Peng Lim, Hwee Hwa Pang

Research Collection School Of Computing and Information Systems

Bipartite graphs can model many real life applications including users-rating-products in online marketplaces, users-clicking-webpages on the World Wide Web and users referring users in social networks. In these graphs, the anomalousness of nodes in one partite often depends on that of their connected nodes in the other partite. Previous studies have shown that this dependency can be positive (the anomalousness of a node in one partite increases or decreases along with that of its connected nodes in the other partite) or negative (the anomalousness of a node in one partite rises or falls in opposite direction to that of its …

Cost-Sensitive Online Classification, Jialei Wang, Peilin Zhao, Steven C. H. Hoi

Research Collection School Of Computing and Information Systems

Both cost-sensitive classification and online learning have been extensively studied in data mining and machine learning communities, respectively. However, very limited study addresses an important intersecting problem, that is, “Cost-Sensitive Online Classification". In this paper, we formally study this problem, and propose a new framework for Cost-Sensitive Online Classification by directly optimizing cost-sensitive measures using online gradient descent techniques. Specifically, we propose two novel cost-sensitive online classification algorithms, which are designed to directly optimize two well-known cost-sensitive measures: (i) maximization of weighted sum of sensitivity and specificity, and (ii) minimization of weighted misclassification cost. We analyze the theoretical bounds of …

Fast And Accurate Psd Matrix Estimation By Row Reduction, Hiroshi Kuwajima, Takashi Washio, Ee Peng Lim

Research Collection School Of Computing and Information Systems

Fast and accurate estimation of missing relations, e.g., similarity, distance and kernel, among objects is now one of the most important techniques required by major data mining tasks, because the missing information of the relations is needed in many applications such as economics, psychology, and social network communities. Though some approaches have been proposed in the last several years, the practical balance between their required computation amount and obtained accuracy are insufficient for some class of the relation estimation. The objective of this paper is to formalize a problem to quickly and efficiently estimate missing relations among objects from the …

Impact Of Multimedia In Sina Weibo: Popularity And Life Span, Xun Zhao, Feida Zhu, Weining Qian, Aoying Zhou

Research Collection School Of Computing and Information Systems

Multimedia contents such as images and videos are widely used in social network sites nowadays. Sina Weibo, a Chinese microblogging service, is one of the first microblog platforms to incorporate multimedia content sharing features. This work provides statistical analysis on how multimedia contents are produced, consumed, and propagated in Sina Weibo. Based on 230 million tweets and 1.8 million user profiles in Sina Weibo, we study the impact of multimedia contents on the popularity of both users and tweets as well as tweet life span. Our preliminary study shows that multimedia tweets dominant pure text ones in SinaWeibo. Multimedia contents …

Divad: A Dynamic And Interactive Visual Analytical Dashboard For Exploring And Analyzing Transport Data, Tin Seong Kam, Ketan Barshikar, Shaun Jun Hua Tan

Research Collection School Of Computing and Information Systems

The advances in location-based data collection technologies such as GPS, RFID etc. and the rapid reduction of their costs provide us with a huge and continuously increasing amount of data about movement of vehicles, people and goods in an urban area. This explosive growth of geospatially-referenced data has far outpaced the planner’s ability to utilize and transform the data into insightful information thus creating an adverse impact on the return on the investment made to collect and manage this data. Addressing this pressing need, we designed and developed DIVAD, a dynamic and interactive visual analytics dashboard to allow city planners …

A Unified Learning Framework For Auto Face Annotation By Mining Web Facial Images, Dayong Wang, Steven C. H. Hoi, Ying He

Research Collection School Of Computing and Information Systems

Auto face annotation plays an important role in many real-world multimedia information and knowledge management systems. Recently there is a surge of research interests in mining weakly-labeled facial images on the internet to tackle this long-standing research challenge in computer vision and image understanding. In this paper, we present a novel unified learning framework for face annotation by mining weakly labeled web facial images through interdisciplinary efforts of combining sparse feature representation, content-based image retrieval, transductive learning and inductive learning techniques. In particular, we first introduce a new search-based face annotation paradigm using transductive learning, and then propose an effective …

Semi-Automatic Simulation Initialization By Mining Structured And Unstructured Data Formats From Local And Web Data Sources, Olcay Sahin

Computational Modeling & Simulation Engineering Theses & Dissertations

Initialization is one of the most important processes for obtaining successful results from a simulation. However, initialization is a challenge when 1) a simulation requires hundreds or even thousands of input parameters or 2) re-initializing the simulation due to different initial conditions or runtime errors. These challenges lead to the modeler spending more time initializing a simulation and may lead to errors due to poor input data.

This thesis proposes two semi-automatic simulation initialization approaches that provide initialization using data mining from structured and unstructured data formats from local and web data sources. First, the System Initialization with Retrieval (SIR) …

Self-Regulating Action Exploration In Reinforcement Learning, Teck-Hou Teng, Ah-Hwee Tan

Research Collection School Of Computing and Information Systems

The basic tenet of a learning process is for an agent to learn for only as much and as long as it is necessary. With reinforcement learning, the learning process is divided between exploration and exploitation. Given the complexity of the problem domain and the randomness of the learning process, the exact duration of the reinforcement learning process can never be known with certainty. Using an inaccurate number of training iterations leads either to the non-convergence or the over-training of the learning agent. This work addresses such issues by proposing a technique to self-regulate the exploration rate and training duration …

Entity Synonyms For Structured Web Search, Tao Cheng, Hady W. Lauw, Stelios Paparizos

Research Collection School Of Computing and Information Systems

Nowadays, there are many queries issued to search engines targeting at finding values from structured data (e.g., movie showtime of a specific location). In such scenarios, there is often a mismatch between the values of structured data (how content creators describe entities) and the web queries (how different users try to retrieve them). Therefore, recognizing the alternative ways people use to reference an entity, is crucial for structured web search. In this paper, we study the problem of automatic generation of entity synonyms over structured data toward closing the gap between users and structured data. We propose an offline, data-driven …

Influentials, Novelty, And Social Contagion: The Viral Power Of Average Friends, Close Communities, And Old News, Nicholas Harrigan, Palakorn Achananuparp, Ee Peng Lim

Research Collection School Of Computing and Information Systems

What is the effect of (1) popular individuals, and (2) community structures on the retransmission of socially contagious behavior? We examine a community of Twitter users over a five month period, operationalizing social contagion as ‘retweeting’, and social structure as the count of subgraphs (small patterns of ties and nodes) between users in the follower/following network. We find that popular individuals act as ‘inefficient hubs’ for social contagion: they have limited attention, are overloaded with inputs, and therefore display limited responsiveness to viral messages. We argue this contradicts the ‘law of the few’ and ‘influentials hypothesis’. We find that community …

Retrieval Of Sub-Pixel-Based Fire Intensity And Its Application For Characterizing Smoke Injection Heights And Fire Weather In North America, David Peterson

Department of Earth and Atmospheric Sciences: Dissertations, Theses, and Student Research

For over two decades, satellite sensors have provided the locations of global fire activity with ever-increasing accuracy. However, the ability to measure fire intensity, know as fire radiative power (FRP), and its potential relationships to meteorology and smoke plume injection heights, are currently limited by the pixel resolution. This dissertation describes the development of a new, sub-pixel-based FRP calculation (FRP_f) for fire pixels detected by the MODerate Resolution Imaging Spectroradiometer (MODIS) fire detection algorithm (Collection 5), which is subsequently applied to several large wildfire events in North America. The methodology inherits an earlier bi-spectral algorithm for retrieving sub-pixel …

The Implementation Of The Shear Correlation Function And The Matter Power Spectrum In R, Allison A. Scheppelmann, Deborah J. Bard

STAR Program Research Presentations

Weak gravitational lensing is an important tool in understanding the large-scale structure of the universe. One component in understanding the effect of weak gravitational lensing is the shear correlation function and matter power spectrum. The calculation of these values is often complicated and time consuming. In order to decrease the cost of these calculations they were implemented in R using parallelization. This resulted in the calculations completing faster and the process to be easily changed in order to fit the need of each researcher using the algorithms created in R.

A Non-Parametric Visual-Sense Model Of Images: Extending The Cluster Hypothesis Beyond Text, Kong-Wah Wan, Ah-Hwee Tan, Joo-Hwee Lim, Liang-Tien Chia

Research Collection School Of Computing and Information Systems

The main challenge of a search engine is to find information that are relevant and appropriate. However, this can become difficult when queries are issued using ambiguous words. Rijsbergen first hypothesized a clustering approach for web pages wherein closely associated pages are treated as a semantic group with the same relevance to the query (Rijsbergen 1979). In this paper, we extend Rijsbergen’s cluster hypothesis to multimedia content such as images. Given a user query, the polysemy in the return image set is related to the many possible meanings of the query. We develop a method to cluster the polysemous images …