Data Science Commons^™

Forecasting The Number Of Monthly Active Facebook And Twitter Worldwide Users Using Arma Model, Qasem Abu Al-Haija, Qian Mao, Kamal Al Nasr

Computer Science Faculty Research

In this study, an Auto-Regressive Moving Average (ARMA) Model with optimal order has been developed to estimate and forecast the short term future numbers of the monthly active Facebook and Twitter worldwide users. In order to pickup the optimal estimation order, we analyzed the model order vs. the corresponding model error in terms of final prediction error. The simulation results showed that the optimal model order to estimate the given Facebook and Twitter time series are ARMA[5, 5] and ARMA[3, 3], respectively, since they correspond to the minimum acceptable prediction error values. Besides, the optimal models recorded a high-level of …

A Grammar For Reproducible And Painless Extract-Transform-Load Operations On Medium Data, Benjamin S. Baumer

Statistical and Data Sciences: Faculty Publications

Many interesting datasets available on the Internet are of a medium size—too big to fit into a personal computer’s memory, but not so large that they would not fit comfortably on its hard disk. In the coming years, datasets of this magnitude will inform vital research in a wide array of application domains. However, due to a variety of constraints they are cumbersome to ingest, wrangle, analyze, and share in a reproducible fashion. These obstructions hamper thorough peer-review and thus disrupt the forward progress of science. We propose a predictable and pipeable framework for R (the state-of-the-art statistical computing environment) …

Exploring And Visualizing Household Electricity Consumption Patterns In Singapore: A Geospatial Analytics Approach, Yong Ying Tan, Tin Seong Kam

Research Collection School Of Computing and Information Systems

Despite being a small country-state, electricity consumption in Singa-pore is said to be non-homogeneous, as exploratory data analysis showed that the distributions of electricity consumption differ across and within administrative boundaries and dwelling types. Local indicators of spatial association (LISA) were calculated for public housing postal codes using June 2016 data to discover local clusters of households based on electricity consumption patterns. A detailed walkthrough of the analytical process is outlined to describe the R packages and framework used in the R environment. The LISA results are visualized on three levels: country level, regional level and planning subzone level. At …

Bridge Deck Delamination Segmentation Based On Aerial Thermography Through Regularized Grayscale Morphological Reconstruction And Gradient Statistics, Chongsheng Cheng, Zhexiong Shang, Zhigang Shen

Department of Construction Engineering and Management: Faculty Publications

Environmental and surface texture-induced temperature variation across the bridge deck is a major source of errors in delamination detection through thermography. This type of external noise poises a significant challenge for conventional quantitative methods such as global thresholding and k-means clustering. An iterative top-down approach is proposed for delamination segmentation based on grayscale morphological reconstruction. A weight-decay function was used to regularize the reconstruction for regional maxima extraction. The mean and coefficient of variation of temperature gradient estimated from delamination boundaries were used for discrimination. The proposed approach was tested on a lab experiment and an in-service bridge deck. The …

Thermographic Laplacian-Pyramid Filtering To Enhance Delamination Detection In Concrete Structure, Chongsheng Cheng, Ri Na, Zhigang Shen

Department of Construction Engineering and Management: Faculty Publications

Despite decades of efforts using thermography to detect delamination in concrete decks, challenges still exist in removing environmental noise from thermal images. The performance of conventional temperature-contrast approaches can be significantly limited by environment-induced non-uniform temperature distribution across imaging spaces. Time-series based methodologies were found robust to spatial temperature non-uniformity but requires extended period to collect data. A new empirical image filtering method is introduced in this paper to enhance the delamination detection using blob detection method that originated from computer vison. The proposed method employs a Laplacian of Gaussian filter to achieve multi-scale detection of abnormal thermal patterns by …

Streaming Feature Grouping And Selection (Sfgs) For Big Data Classification, Noura Helal Hamad Al Nuaimi

Dissertations

Real-time data has always been an essential element for organizations when the quickness of data delivery is critical to their businesses. Today, organizations understand the importance of real-time data analysis to maintain benefits from their generated data. Real-time data analysis is also known as real-time analytics, streaming analytics, real-time streaming analytics, and event processing. Stream processing is the key to getting results in real-time. It allows us to process the data stream in real-time as it arrives. The concept of streaming data means the data are generated dynamically, and the full stream is unknown or even infinite. This data becomes …

Untapped Potential Of Clinical Text For Opioid Surveillance, Amy L. Olex, Tamas Gal, Majid Afshar, Dmitriy Dligach, Niranjan Karnik, Travis Oakes, Brihat Sharma, Meng Xie, Bridget T. Mcinnes, Julian Solway, Abel Kho, William Cramer, F. Gerard Moeller

Wright Center for Clinical and Translational Research Works

Accurate surveillance is needed to combat the growing opioid epidemic. To investigate the potential volume of missed opioid overdoses, we compare overdose encounters identified by ICD-10-CM codes and an NLP pipeline from two different medical systems. Our results show that the NLP pipeline identified a larger percentage of OOD encounters than ICD-10-CM codes. Thus, incorporating sophisticated NLP techniques into current diagnostic methods has the potential to improve surveillance on the incidence of opioid overdoses.

Radically Simplifying Gated Recurrent Architectures Without Loss Of Performance, Jonathan Boardman, Ying Xie

Published and Grey Literature from PhD Candidates

Long Short-Term Memory (LSTM) units are a family of Recurrent Neural Network (RNN) architectures that have proven incredibly effective at learning from sequence data. They are also extremely complex, making them expensive to train and difficult to understand. A recent trend towards simplification has produced the Gated Recurrent Unit (GRU) and the Minimal Gated Unit (MGU), both of which perform as well as the LSTM (or better) on a variety of tasks. The MGU is one of the simplest gated recurrent architectures at the moment. Our study demonstrates that it is possible to radically simplify the MGU without significant loss …

Canadian Hockey Leagues Game-To-Game Performance, Nick R. Riccardi

Sport Management - All Scholarship

This study examines game-to-game performance of players across the three Canadian Hockey Leagues (Western Hockey League, Ontario Hockey League, and Quebec Major Junior Hockey League) for the 2017-2018 season. It tests the importance of factors such as rest, travel, weather conditions, and more. Data for this study were collected from each of the three CHL websites and from www.weatherunderground.com. The null hypotheses of different factors affecting performance were tested through regression models using Ordinary Least Squares. The dependent variables, used across different specifications, were on-ice performance variables such as points, goals, and penalty minutes on a per-game basis.

On The Inability Of Markov Models To Capture Criticality In Human Mobility, Vaibhav Klukarni, Abhijit Mahalunkar, Benoit Garbinato, John Kelleher

Conference papers

We examine the non-Markovian nature of human mobility by exposing the inability of Markov models to capture criticality in human mobility. In particular, the assumed Markovian nature of mobility was used to establish an upper bound on the predictability of human mobility, based on the temporal entropy. Since its inception, this bound has been widely used for validating the performance of mobility prediction models. We show that the variants of recurrent neural network architectures can achieve significantly higher prediction accuracy surpassing this upper bound. The central objective of our work is to show that human-mobility dynamics exhibit criticality characteristics which …

Enrollment And Assessment Of A First-Year College Class Social Network For A Controlled Trial Of The Indirect Effect Of A Brief Motivational Intervention, Nancy P. Barnett, Melissa A. Clark, Shannon R. Kenney, Graham Diguiseppi, Matthew K. Meisel, Sara Balestrieri, Miles Q. Ott, John Light

Statistical and Data Sciences: Faculty Publications

Heavy drinking and its consequences among college students represent a serious public health problem, and peer social networks are a robust predictor of drinking-related risk behaviors. In a recent trial, we administered a Brief Motivational Intervention (BMI) to a small number of first-year college students to assess the indirect effects of the intervention on peers not receiving the intervention. Objectives: To present the research design, describe the methods used to successfully enroll a high proportion of a first-year college class network, and document participant characteristics. Methods: Prior to study enrollment, we consulted with a student advisory group and campus stakeholders …

Data Sharing At Scale: A Heuristic For Affirming Data Cultures, Lindsay Poirier, Brandon Costelloe-Kuehn

Statistical and Data Sciences: Faculty Publications

Addressing the most pressing contemporary social, environmental, and technological challenges will require integrating insights and sharing data across disciplines, geographies, and cultures. Strengthening international data sharing networks will not only demand advancing technical, legal, and logistical infrastructure for publishing data in open, accessible formats; it will also require recognizing, respecting, and learning to work across diverse data cultures. This essay introduces a heuristic for pursuing richer characterizations of the “data cultures” at play in international, interdisciplinary data sharing. The heuristic prompts cultural analysts to query the contexts of data sharing for a particular discipline, institution, geography, or project at seven …

Classification As Catachresis: Double Binds Of Representing Difference With Semiotic Infrastructure, Lindsay Poirier

Statistical and Data Sciences: Faculty Publications

Background; This article explores the results of a three-year ethnographic study of how semiotic infrastructures-or digital standards and frameworks such as taxonomies, schemas, and ontologies that encode the meaning of data-are designed. Analysis: It examines debates over best practices in semiotic infrastructure design, such as how much complexity adopted languages should characterize versus how restrictive they should be. It also discusses political and pragmatic considerations that impact what and how information is represented in an information system. Conclusion and implications: This article suggests that all databased representations are forms of data power, and that examining semiotic infrastructure design provides insight …

Explorobot: Rapid Exploration With Chart Automation, Tamara Matthews, Rohan Goel, John Mcauley

Conference papers

General-purpose visualization tools are used by people with varying degrees of data literacy. Often the user is not a professional analyst or data scientist and uses the tool infrequently, to support an aspect of their job. This can present difficulties as the user’s unfamiliarity with visualization practice and infrequent use of the tool can result in long processing time, inaccurate data representations or inappropriate visual encodings. To address this problem, we developed a visual analytics application called exploroBOT. The exploroBOT automatically generates visualizations and the exploration guidance path (an associated network of decision points, mapping nodes where visualizations change). These …

Deep Patient Representation Of Clinical Notes Via Multi-Task Learning For Mortality Prediction, Yuqi Si, Kirk Roberts

Faculty, Staff and Student Publications

We propose a deep learning-based multi-task learning (MTL) architecture focusing on patient mortality predictions from clinical notes. The MTL framework enables the model to learn a patient representation that generalizes to a variety of clinical prediction tasks. Moreover, we demonstrate how MTL enables small but consistent gains on a single classification task (e.g., in-hospital mortality prediction) simply by incorporating related tasks (e.g., 30-day and 1-year mortality prediction) into the MTL framework. To accomplish this, we utilize a multi-level Convolutional Neural Network (CNN) associated with a MTL loss component. The model is evaluated with 3, 5, and 20 tasks and is …

Knowledge Management Overview Of Feature Selection Problem In High-Dimensional Financial Data: Cooperative Co-Evolution And Map Reduce Perspectives, A. N. M. Bazlur Rashid, Tonmoy Choudhury

Research outputs 2014 to 2021

The term "big data" characterizes the massive amounts of data generation by the advanced technologies in different domains using 4Vs volume, velocity, variety, and veracity-to indicate the amount of data that can only be processed via computationally intensive analysis, the speed of their creation, the different types of data, and their accuracy. High-dimensional financial data, such as time-series and space-Time data, contain a large number of features (variables) while having a small number of samples, which are used to measure various real-Time business situations for financial organizations. Such datasets are normally noisy, and complex correlations may exist between their features, …

Population Data Centre Profile - The Western Australian Data Linkage Branch, Steve Hodges, Tom Eitelhuber, Alexandra Merchant, Janine Alan

Research outputs 2014 to 2021

Established in 1995, the Western Australian Data Linkage Branch (DLB) is Australia’s longest running data linkage agency. The Western Australian Data Linkage System (WADLS) employs an enduring linkage model spanning over 60 data collections supported by internally developed and supported software and IT infrastructure. DLB has delivered, and continues to deliver, a range of significant data linkage innovations, many of which have been adopted elsewhere. A current restructure within the Western Australian Department of Health (which we will refer to as the Department of Health) will provide an improved funding model geared toward addressing issues with staff retention, capacity and …

Detecting Special-Cause Variation 'Events' From Process Data Signatures, Timothy M. Young, Olga Khaliukova, Nicolas André, Alexander Petutschnigg, Timothy G. Rials, Chung-Hao Chen

Electrical & Computer Engineering Faculty Publications

The ability to detect the special-cause variation of incoming feedstocks from advanced sensor technology is invaluable to manufacturers. Many on-line sensors produce data signatures that require further off-line statistical processing for interpretation by operational personnel. However, early detection of changes in variation in incoming feedstocks may be imperative to promote early-stage preventive measures. A method is proposed in this applied study for developing control bands to quantify the variation of data signatures in the context of statistical process control (SPC). Control bands based on pointwise prediction intervals constructed from the Bonferroni Inequality and Bayesian smoothing splines are developed. Applications using …

The Global Disinformation Order: 2019 Global Inventory Of Organised Social Media Manipulation, Samantha Bradshaw, Philip N. Howard

Copyright, Fair Use, Scholarly Communication, etc.

Executive Summary

Over the past three years, we have monitored the global organization of social media manipulation by governments and political parties. Our 2019 report analyses the trends of computational propaganda and the evolving tools, capacities, strategies, and resources.

1. Evidence of organized social media manipulation campaigns which have taken place in 70 countries, up from 48 countries in 2018 and 28 countries in 2017. In each country, there is at least one political party or government agency using social media to shape public attitudes domestically.

2.Social media has become co-opted by many authoritarian regimes. In 26 countries, computational propaganda …