Physical Sciences and Mathematics Commons^™

Text Anomaly Detection With Arae-Anogan, Tec Yan Yap

Honors Projects

Generative adversarial networks (GANs) are now one of the key techniques for detecting anomalies in images, yielding remarkable results. Applying similar methods to discrete structures, such as text sequences, is still largely an unknown. In this work, we introduce a new GAN-based text anomaly detection method, called ARAE-AnoGAN, that trains an adversarially regularized autoencoder (ARAE) to reconstruct normal sentences and detects anomalies via a combined anomaly score based on the building blocks of ARAE. Finally, we present experimental results demonstrating the effectiveness of ARAE-AnoGAN and other deep learning methods in text anomaly detection.

A Data Analysis Of The World Happiness Index And Its Relation To The North-South Divide, Charles Alba

Undergraduate Economic Review

In this document, we perform a detailed data analysis on the World Happiness Report with its relation to the socio-economic North-South Divide. In order to do so, we perform some extensive data cleaning and analysis before querying on the World Happiness Report. Our results based on Hypothesis Testing determines the happiness of the Global North is greater than that of the Global South. Furthermore, our queries show that the mean happiness score for the Global North significantly outweighing that of the South. Likewise, the 10 'Happiest' nations all belong to the Global North whereas the 10 'least happy' nations belong …

Encoding Lexicographical Ordering Constraints In Sat, Wenting Zhao

Honors Projects

Symmetry occurs in many constraint satisfaction problems, and it is important to deal with it efficiently and effectively, as it often leads to an exponential number of isomorphic assignments. Symmetric rows and columns in matrices are an important class of symmetries in constraint programming. In this work, we develop a new SAT encoding for partial lexicographical ordering constraints to break symmetries in such places. We also survey all the previous complete lex-leader encodings in literature and translate them into SAT encodings. We perform experimental analysis on how these lex-leader constraints impact the solving of Balanced Incomplete Block Design (BIBD) instances. …

2-(51, 6, 1) Block Designs, Wenting Zhao, Mark Liffiton. Faculty Advisor

John Wesley Powell Student Research Conference

No abstract provided.

Parallel Infeasibility Analysis, Wenting Zhao, Mark Liffiton, Faculty Advisor

John Wesley Powell Student Research Conference

Oral presentation abstract.

Tabulating Pseudoprimes And Tabulating Liars, Andrew Shallue

Scholarship

This paper explores the asymptotic complexity of two problems related to the Miller-Rabin-Selfridge primality test. The first problem is to tabulate strong pseudoprimes to a single fixed base $a$. It is now proven that tabulating up to $x$ requires $O(x)$ arithmetic operations and $O(x\log{x})$ bits of space.The second problem is to find all strong liars and witnesses, given a fixed odd composite $n$.This appears to be unstudied, and a randomized algorithm is presented that requires an expected $O((\log{n})^2 + |S(n)|)$ operations (here $S(n)$ is the set of strong liars).Although interesting in their own right, a notable application is the search …

Contact Angle Measurement, Wenting Zhao, Mark Liffiton, Faculty Advisor

John Wesley Powell Student Research Conference

The Contact angle, where a liquid/vapor interface meets a solid surface[wiki], has been widely used to measure the wettability of a surface in physics and chemistry. Scientists place a drop on a surface of interest, take an image of the drop in profile, and measure the angle the drop makes with the surface. We have developed a Contact Angle Measurement plugin for the ImageJ image analysis framework, which provides researchers a easier way to access experiment data. The major goal of our algorithm is to automatically detect drops and surfaces via image analysis, so that we can calculate the contact …

Simmons' Research Aims To Improve Function For Online Security, Danielle Kamp

News and Events

No abstract provided.

Computer Science Major Interns At Google, Danielle Kamp

News and Events

No abstract provided.

Dougan Award's Sponsor Named Recipient, Kim Hill

News and Events

No abstract provided.

Constructing Carmichael Numbers Through Improved Subset-Product Algorithms, W.R. Alford, Jon Grantham, Steven Hayman, Andrew Shallue

Scholarship

We have constructed a Carmichael number with 10,333,229,505 prime factors, and have also constructed Carmichael numbers with prime factors for every between 3 and 19,565,220. These computations are the product of implementations of two new algorithms for the subset product problem that exploit the non-uniform distribution of primes with the property that divides a highly composite .

Elitist Schema Overlays: A Multi-Parent Genetic Operator, Nick Nichols, Mark Liffiton, Faculty Advisor

John Wesley Powell Student Research Conference

Genetic Algorithms are programs inspired by natural evolution used to solve difficult problems in Mathematics and Computer Science. The theoretical foundations of Genetic Algorithms, the schema theorem and the building-block hypothesis, state that the success of Genetic Algorithms stems from the propagation of fit genetic subsequences. Multi-parent operators were shown to increase the performance of Genetic Algorithms by increasing the disruptivity of genetic operations. Disruptive genetic operators help prevent suboptimal genetic sequences from propagating into future generations, which leads to an improved fitness for the population over time. In this paper we explore the use of a novel multi-parent genetic …

Liffiton Develops Free App For Teachers, Kim Hill

News and Events

No abstract provided.

Analyzing And Extending An Infeasibility Analysis Algorithm, Ammar Malik

Honors Projects

Constraint satisfaction problems (CSPs) involve finding assignments to a set of variables that satisfy some mathematical constraints. Unsatisfiable constraint problems are CSPs with no solution. However, useful characteristic subsets of these problems may be extracted with algorithms such as the MARCO algorithm, which outperforms the best known algorithms in the literature. A heuristic choice in the algorithm affects how it traverses the search space to output these subsets. This work analyzes the effect of this choice and introduces three improvements to the algorithm. The first of these improvements sacrifices completeness in terms of one type of subset in order to …

Susan Anderson-Freed, Susan Anderson-Freed, Meg Miner

All oral histories

Dr. Susan Anderson-Freed began teaching in the Sociology Department at Illinois Wesleyan University in 1977. A few years later she was reappointed to a position in the Computer Science program where she remained until her departure in 2008.

Anderson-Freed left teaching due to the effects of cancer treatments she was receiving and died on November 4, 2012. A long-time knitting enthusiast, she used the time of her treatments to design patterns and published two books on the subjects. Proceeds from these volumes, which have already been translated into several languages, are designated for the Community Cancer Center, Normal, Illinois. Her …

Native Cardinality Constraints: More Expressive, More Efficient Constraints, Jordyn C. Maglalang

Honors Projects

Boolean cardinality constraints are commonly translated (encoded) into Boolean CNF, a standard form for Boolean satisfiability problems, which can be solved using a standard SAT solving program. However, cardinality constraints are a simple generalization of clauses, and the complexity entailed by encoding them into CNF can be avoided by reasoning about cardinality constraints natively within a SAT solver. In this work, we compare the performance of two forms of native cardinality constraints against some of the best performing encodings from the literature. We designed a number of experiments, modeling the general use of cardinality constraints including crafted, random and application …

Monte Carlo Simulations Of Electron Scattering Experiments, Alan Russian, Bruno Deharak, Faculty Advisor, Mark Liffiton, Faculty Advisor

John Wesley Powell Student Research Conference

This project aims to look at the impact made by certain approximations in electron scattering experiments—specifically whether accounting for these approximation errors is necessary. When using a moveable gun mount, the interaction volume can be determined using a line and cylinder approximation. Data is presented comparing this approximation to the actual volume computed using a Monte Carlo method. A uniform gas distribution is compared to a cosine-squared distribution gas distribution. Additionally, an energy spectrum of a uniform beam is compared to a Gaussian beam for various polarization angles, and a comparison is made between representing the beam as a cone …

Optimization And Analysis Of A Robotic Navigational Algorithm, Derek Carlson, Joshua Brown Kramer, Faculty Advisor

John Wesley Powell Student Research Conference

The problem of robot navigation involves planning a path to move a robot from a start point to a known target point within an obstacle course. The efficiency of such an algorithm can be measured in several ways. For instance, Lumelsky and Stepanov measure the length of the path taken in terms of obstacle perimeters. Gabriely and Rimon compare their two-dimensional algorithm's efficiency to that of the optimal algorithm. Brown Kramer and Sabalka expand upon the work of Gabriely and Rimon to produce an algorithm for dimensions greater than two. The primary objective of this research was to implement improvements …

An Implementation Of A Path Planning Algorithm For A Robot, Maia Czwornog, Joshua Brown Kramer, Faculty Advisor

John Wesley Powell Student Research Conference

No abstract provided.

M-Refinable Extensions Of Real Valued Functions, John Meuser, Tian-Xiao He, Faculty Advisor

John Wesley Powell Student Research Conference

No abstract provided.

Sascha Vitzhum, Jim Browne

Interviews for WGLT

Jim Browne interviews Assistant Professor Sascha Vitzhum about virtual worlds, economics and marketing. (requires RealPlayer)

Student Rights Advocate And Author To Speak, Teresa A. Sherman

News and Events

No abstract provided.

Rapid Face Detection Using Independent Component Analysis, Aditya Rajgarhia '07

Honors Projects

Face detection is the task of determining the locations and sizes of human faces in arbitrary digital images, while ignoring any other objects to the greatest possible extent. A fundamental problem in computer vision, it has important applications in fields ranging from surveillance-based security to autonomous vehicle navigation. Although face detection has been studied for almost a decade, the results are not satisfactory for a variety of practical applications, and the topic continues to receive attention.

A commonly used approach for detecting faces is based on the techniques of "boosting" and "cascading", which allow for real-time face detection. However, systems …

Unsupervised Learning To Improve Anomaly Detection, Daniel H. Garrette '06

Honors Projects

An intrusion detection system (IDS) is used to determine when a computer or computer network is under attack. Most contemporary IDSs operate by defining what an intrusion looks like and checking traffic for matching patterns in network traffic. This approach has unavoidable limitations including the inability to detect novel attacks and the maintenance of a rule bank that must grow with every new intrusion discovered. An anomaly detection scheme attempts to define what is normal so that abnormal traffic can be distinguished from it. This thesis explores the ways that an unsupervised technique called "clustering" can be used to distinguish …

Limits Of Diagonalization And The Polynomial Hierarchy, Kyle Barkmeier '06

Honors Projects

Determining the computational complexity of problems is a large area of study. It seeks to separate these problems into ones with "efficient" solutions, and those with "inefficient" solutions. Of course, the strata is much more fine-grain than this. Of special interest are two classes of problems: P and NP. These have been of much interest to complexity theorists for quite some time, because both contain many instances of important real-world problems, and finding efficient solutions for those in NP would be beneficial for computing applications. Yet with all this attention, there are still important unanswered questions about the two classes. …

Using Binary Space Subdivision To Optimize Primary Ray Processing In Ray-Tracing Algorithms, Mark Portolese '05

Honors Projects

Ray-tracing algorithms have the potential to create extremely realistic three-dimensional computer graphics. The basic idea is to trace light rays from the user through the computer screen into the hypothetical three-dimensional world. This is done to determine what objects should be displayed on the screen. Furthermore, these rays are traced back to the light sources themselves to determine shading and other photorealistic effects. However, without optimization these algorithms are slow and impractical. This paper explores the use of the classic binary space subdivision algorithm in order to speed up the process. Binary space subdivision is the use of binary trees …

Automated Annotation Of Heegaard Diagrams, Dmitry Mogilevsky '03

Honors Projects

No abstract provided.

P-Fastus: Information Extraction System Implemented In A Constraint Programming Language -Sicstus Prolog, Rajen Subba '03

Honors Projects

P-FASTUS is an Information Extraction system developed in SICStus Prolog based on the implementation of FASTUS. It is program that extracts prespecified information such as the name of the companny, location and the position being advertised from" Job PostingIs'' in text files. The system is composed of different levels of processing phases that are implemented using finite-state transducers.

The Virtual Beta: An Interactive Fish Using Java Script And Css, Lauren B. Carroll '03

Honors Projects

My first version of the fish was written in C, because it was the language I was studying at the time. I wanted something that more people could view easily, however and one that could manage images more simply than C. I ended up choosing ]avaScript. At first I tried to translate my C program directly to ]avaScript, but I soon found this was too complex a task.

Narendra Jaggi, Charlie Schlenker

Interviews for WGLT

Charlie Schlenker interviews Professor of Physics Narendra Jaggi about his work with computer modeling of ethnic living choices. (requires RealPlayer)