The Q-Club, which is short for "Quantitative Club," is the departmental student organization, boasting over fifty members. Students speak on their research or share their internship and summer program experiences. Scheduled events take place approximately every other Thursday 11:50 am - 12:20 pm with pizza and beverages being served. Q-Club will meet in Valentine Hall, Room #205/6.
Faculty members occasionally give talks as well, on topics ranging from "Math and Horror" to "An Outrageously Brief History of Mathematics."
If you know of students interested in giving a talk during the semester, please contact Ivan Ramler. (Q-Club Archive Page)
Schedule is as follows:
Thursday, December 5, 2019 at 11:50 am - 12:20 pm in Valentine 205-206.
Speaker: Shuyu Qiao
Title: Using MCMC and Graph Theory to Investigate Diversity of Birds Species
Abstract: We look at methods to access competitions among bird species based on Diamond’s finch data. The data matrix has rows for 13 different finch species and columns corresponding to 17 islands in Vanuatu. In each cell, a 1 represents the presence of the species on that island, and 0 is the absence. If the dot product for the rows for two species is zero, we may have evidence of competition. We call this a checkboard, 1s for one species are always matched with 0s for the other species. How do we know when the number of checkboards is unusually high, more than we would expect by random chance indicating competition? We investigate this question by simulating lots of random matrices and counting the number of checkerboards. To generate the random matrices, we use a Markov Chain to move from one matrix to the next. The connections between the matrices form the edges of a graph, and we use the graph to calculate a transition matrix and investigate the behavior of the MCMC.
Thursday, November 14, 2019 at 11:50 am - 12:20 pm in Valentine 205-206.
Speaker: Sarah Coburn from Career Services - What they can do for you, what our majors have done in the past. How to get in touch with alumni, etc. regarding both full time employment and internship opportunities.
Thursday, November 7, 2019 at 11:50 am - 12:20 pm in Valentine 205/206.
Speaker: Alan Mersereau '66
Title: "Mathematics in Business"
Abstract: I finished my career in in corporate financial plans and controls. Major responsibilities included budgeting, forecasting the income statement and projected cash flow. I was also heavily involved in streamlining our inter functional business processes including customer order fulfillment, and field service logistics. Earlier in my career, I was in charge of the implementation of business applications including Environment Resource Planning, and Just-in-Time operations planning. I will discuss the use of mathematics in my business career.
Thursday, October 31, 2019 at 11:50 am - 12:20 pm in Valentine 205/206.
Speaker: Dr. Michael Sheard
Title: "A Derivation of My Favorite Formula"
I will present a derivation of Euler's original solution to the Basel Problem, which shows that a simple infinite series adds up to the unexpected sum of "pi 2 / 6". This is a fun and surprising application of some basic algebra and calculus.
Thursday, October 17, 2019 at 11:50 am - 12:20 pm in Valentine 205/206.
Speaker, Susan Liu
Title: "Using Cluster Analysis to Understand Long Term Viability of Cards in Clash Royale".
Esports, a multiplayer video game played competitively for spectators, have become increasingly popular and mainstream among professional and casual players today. Clash Royale, a real time multiplayer game that combines elements from collectible card games, tower defense, and multiplayer online battle arena has reached $1 billion in revenue in less than a year on the market (Takahashi, 2017). Players build decks of eight cards which are used to defeat their opponent. Fan sites offer card win rates to assist players in identifying viable cards.
For example, the site Statsroyale, reports cards’ win rates along with their popularity. However, factors such as player skill level (measured in trophies) can skew the overall win rates for cards. This project aims to collect Clash Royale data from Kaggle and Clash Royale application programming interface (API), visualize card performance across trophy count, and identify cards with long term viability using cluster analysis.
Thursday, September 26, 2019 at 11:50 am - 12:20 pm in Valentine 205/206.
Speaker: Xin Tao
Title: Build an All-Voice Reminder
Reminder systems, which allow people to set their devices to remind them to do things, can be helpful to people’s daily lives and increase their efficiency. For some elders, they become forgetful about daily things such as taking medicine; automated reminders could help them remember important tasks. For some students, time management is a challenge; using reminder systems, they could keep themselves on track. However, reminder systems also have limitations. Elders may find them difficult to use if they struggle using computers and other devices in general. Some students can be so distracted by social media and computer games that they ignore their reminders, and as a result, they do not follow the plans they make. The purpose of my research is to build a reminder system that operates entirely by voice.
Thursday, September 12, 2019 at 11:50am-12:20pm in Valentine 205/206.
Speaker: Sai Wei
Title: Aptapp - An In-app Program that Helps Novice User with WeChat Functions
Abstract:WeChat is the most widely used, multi-function super app in China, bringing convenience to every aspect of users’ daily life. However, novice users, those who are new to technology, are having a hard time trying to catch up with the updating functions in WeChat. Aptapp is a WeChat mini-program that allows users to easily get access to specific WeChat guidance with an intelligent customer service responding system. Launched by WeChat team in 2017, mini-programs are the sub-applications aiming to provide more possibilities for WeChat users. Customer service is one of the function components supported by the mini-program API. Working with cloud functions, the customer service in Aptapp is able to respond to users’ requests by sending back the corresponding web-link to the requested guidance. In order to make Aptapp more helpful for novice users, by analyzing the updating features and API’s from WeChat, Aptapp is expected to require less interactions but to provide more intelligent responding system to users in the future.
Schedule is as follows:
Thu, Apr. 25 - Eliza Oliver (Patti Lock)
Mon, Apr. 29 @ 4pm - A Clarkson professor on Quantum Computing (Lisa Torrey)
Speaker: Eliza Oliver, Thursday, April 25, 11:50 am - 12:20 pm in Valentine 205/6
Title: Using Game Theory to Examine the Effects of Prejudice
Abstract: While prejudice and hatred continue to permeate the American public through media and prominent public figures, the need for unity has never been greater. Understanding the effects of intentional and inadvertent prejudice is the first step toward finding a solution for this unnecessary conflict. The Prisoner’s Dilemma studies strategies for iterative decision making between two parties. It has applications to business and economics, psychology, and international relations. Through the discussion of various strategies players can employ in the Prisoner’s Dilemma, we can see how cooperation with others and defection against can enhance the effects of prejudice in a society.Focusing on the most prominent strategies, including Tit-for-Tat, and considering various conditions of the environment, we will be able to further understand prejudice and discuss efforts that can be made to further reduce the spread of prejudice.
Speaker: Zeyu Lu, Thursday, April 11, 11:50 am - 12:20 pm in Valentine 205/6
Title: Knot Theory, the Jones polynomial, and graph theory
Abstract: Knot theory is the study of mathematical knots. A main focus is whether two knots are the same even though they may appear different. Knot invariants play an important role in answering this question. One of the most interesting and successful types of invariants are knot polynomials. These involve associating polynomials with knots. We investigated several knot polynomials, including the Jones polynomial, as well as how we can relate these to graph theory
Speaker: Allie Withee, Thursday, March 28, 11:50 - 12:20 in Valentine 205/6.
Title: A Shiny App for Visualizing Adirondack River Data
Abstract: A Shiny Application within R was created to aid geologists in the visualization and time series analysis of Adirondack River discharge.
The data include 23 rivers with their monthly discharge levels starting in 1908 and ending in 2018. The goal is to find trends in the data to understand historical patterns and explore the effects of climate change. Analysis includes simple exponential smoothing, seasonal and simple linear regression models, Holt Winters, and a monthly spaghetti plot to highlight key measures and trends in the data.
Speakers: Guinevere Gilman, Angelica Munyao, Matthijs van Mierlo, Thursday, March 14, 11:50 am - 12:20 pm in Valentine 205/6.
Title: Synesthesia: Sensor Visualization Android Application
Abstract:Synesthesia is a perceptual phenomenon in which the stimulation of one sensory organ results in an experience in a different sensory organ.
Our Android application allows a user to create images using touch input and the manipulation of sensor data. The sensors used are the gyroscope to gather angular velocity data, the microphone for sound input, and the light sensor for brightness. With the aid of an audio-processing Java library known as TarsosDSP, our application simulates a somewhat synesthetic experience in translating environmental data to a visual representation.
Speaker: Travis Marnell, Thursday, 2/28/19, 11:50 am - 12:20 pm in Valentine Hall, Room #205/6.
Title : How Math is Solving the Issue of Gerrymandering
Abstract: In the current political sphere, there are many injustices and hidden dilemmas that are plaguing our government. One is in the realm of fair division, and this is where we run into the problem of gerrymandering.
Gerrymandering is an issue that arises when states unfairly divide their congressional districts. A district must fit within three criteria, equal populations, contiguous and compact. The compactness of each district is the main complication when it comes to mappings.
The main measurement of compactness is The Reock measure, which is equal to area of the district over the area of the smallest enclosing circle. By looking into this, a case has been brought up in Pennsylvania where the prosecutors based most of their evidence around the statistics that come with The Reock measure as well as other math data. A similar case is coming to the supreme court in late March where North Carolina, Maryland and Wisconsin are looking to be redistricted based on these measurements. Focusing on these cases and some other sources we can see how math and statistics is helping solve some of the unfairness that is found in our government system.
Speaker: Emily Viehl, Tuesday, 1/29/19, 11:50 am - 12:20 pm, in Valentine Hall, Room #205/6.
Title: Creating a Book Recommendation System for Project Gutenberg
Abstract: Project Gutenberg is a free online library that contains over 56,000 eBooks that are available to be downloaded. Like many other online repositories, the number of eBooks available on Project Gutenberg has increased dramatically over the last ten years and is continuing to grow. What makes Project Gutenberg unique is that in addition to the books being freely available on Kindle, they can also be downloaded into R.The goal of this project was to create a recommendation system for the fiction books found on Project Gutenberg using cluster analysis. Recommendation systems are often based on user patterns/feedback however, in the absence of user data, recommendations can be made based on content. This particular recommendation system makes suggestions based on sentiment, style, and content.
We will have our last Fall semester Q-Club meeting from 12:00 pm-1:00 pm on Friday, December 14 in Bloomer Auditorium (Brown Hall #122).
Dr. Brian Macdonald, Principal Consultant, GreaterThanPlusMinus, LLC Adjunct Professor, Department of Management Science, University of Miami Adjunct Professor, Sports Analytics, Florida Atlantic University College of Business
Title: "Using analytics for league realignment and player evaluation in sports."
Abstract: We give examples of how data visualization, analysis, and optimization can be used within the National Hockey League, and in the sports industry in general, in a variety of different contexts. We first present an example of how optimization techniques can be used at the league level to help make decisions about realignment when a team relocates, or when a new team is added to the league. We also discuss how analytics can be used to assist a team’s front office, coaching staff, and scouting department when they make personnel or coaching decisions.
Yuexin Li, Tuesday, December 4, 11:50 am - 12:20 pm in Bloomer Auditorium (Brown Hall #122)
Title: "Improve Computer’s Ability to Read Texts from Text Images with Complex Layouts Using Hierarchical Clustering and Smoothed N-gram Model".
Abstract: As artificial intelligence develops rapidly these years, people hope to teach computers the ability to read texts from paper so that paper-based material could be digitalized more efficiently for further use in various machine processes such as machine translation, text-to-speech, data mining etc. Tesseract-ocr, the text extraction tool behind Google photo translate service, helps people extract texts from images of texts, however it has limitations when dealing with complex layouts, such as multi-columns.
This research tried to improve Tesseract-OCR’s performance of images with non-uniformed layouts, using two artificial intelligence approaches: hierarchical clustering and smoothed n-gram models. When using hierarchical clustering, I interrupted Tesseract-OCR’s process to perform better block segmentation by grouping texts together from character level. For n-gram models, I directly worked on the extracted raw strings and tried to identify and fix miss-ordered text blocks based on word frequency
Tiara Davis from Career Services, Tuesday, 10/30/18, 11:50 am - 12:20 pm in Bloomer Auditorium (Brown Hall #122)
What Career Services can do for you, what our majors have done in the past. How to get in touch with alumni, etc. regarding both full time employment and internship opportunities.
Seongwon (Ryan) Im Tuesday, 10/16/18, 11:50 am - 12:20 pm in Bloomer Auditorium (Brown Hall #122)
Title: "Design-weighted Regression Adjusted Plus-Minus (D-WRAP-M) for Evaluation of Player Impact"
Abstract: Ice hockey is one example of fluid team sports. Not only do twelve players, six per team, with one puck participate in a play simultaneously, but players from each team also come on and off the ice repeatedly. This presents challenges for evaluating the impact of individual players. Traditionally, the game statistics such as goals, shots, and passes tend to be the typical criteria for the evaluation. However, incapability to take account for the impact of all the players involved in plays rose as the main drawback of such method.
Regression Adjusted Plus-Minus (RAPM) is one widely accepted way to overcome the drawback. RAPM builds a design matrix taking players presence using +1/0/-1 to estimate the player performance. During summer research, we have taken a further step, not only building a design matrix using players’ presence but also taking account of players’ distances from events to better estimate the players’ performance.
Anna Breton, Allison Reiner, Kenneth Schimpf 10/2/18, Tuesday, 11:50 am-12:20 pm in Bloomer Auditorium (Brown Hall #122)
Title: "Summer Internship and Research Experience"
Abstract:Anna Breton interned at Esri with a team that creates cartography extensions for Adobe Creative Cloud. She also created data visualization prototypes for their cartography software.
Allison Reiner received a fellowship to conduct academic research, and performed biostatistical analysis to determine how biological sex affects gene expression in the eye (used the cornea tissue of mice).
Kenneth Schimpf went to the Cyber Security REU program at Wright State University. The research project involved reveres engineering M7700 ECU binaries to syntactically identify sensor addresses in read-only memory.
Guinevere Gilman 09/18/18, Tuesday 12:00 pm- 12:30 pm in Bloomer Auditorium (Brown 122)
Title: "High Bandwidth Mode in Overwatch"
Abstract: Interactive applications such as multiplayer online games require user interactions to be possible, because handled responsively, with the smallest amount of delay users must respond rapidly to the actions of other users in these types of games. A delay of even 50ms has been determined to make some multiplayer games to be frustrating and unplayable in previous investigations. However, there has not been sufficient research conducted to determine an acceptable threshold of delay for the playability of more modern games, as much of the data collected has been from games created in the 1990s and early 2000s.
This research focused on one of the most popular modern multiplayer team-based multiplayer games, known as Overwatch, which has made the claim that its relatively new high bandwidth mode will drastically increase the quality of experience bandwidth. It for players. It also claims to group is a first-person shooter, where two players with teams of six similar players compete against each other to fulfill various objectives using player characters with different types of weapons including hit-scan and projectile-based ones.
Lara Clemens 09/08/18, Thursday in Bloomer Auditorium (Brown 122)
Title: “A mathematical modeling approach to behaviors of multiply-phosphorylated intrinsically disordered proteins in signaling networks”
Abstract: Immune cells detect invaders and trigger a defensive response via a network of protein interactions both outside and inside the cell. Classically, the interactions of these proteins are understood by studying the structure of each protein. However, some proteins involved in the immune response have intrinsically disordered regions, including some with no cytoplasmic structure at all. Despite this, these unstructured signaling molecules have been found to exhibit complex nonlinear behavior, including cooperativity and sequential binding. For example, the intrinsically disordered T Cell Receptor zeta chain has six tyrosines and exhibits an approximately 100-fold rate enhancement from first to sixth phosphorylations. Here, we computationally model the zeta chain with a freely-jointed chain model coupled to idealized spherical binding enzymes. Following experimental and theoretical evidence, we explore three model assumptions: (1) phosphorylation stiffens nearby peptide bonds, (2) phosphorylation changes the electrostatic interaction of the chain with the inner leaflet of the membrane, and (3) phosphorylation allows enzymes to remain bound to the phospho-tyrosine. We find that, first, entropic flexibility alone leads to cooperativity or anti-cooperativity in kinase activity, and, second, that sequential binding emerges naturally from entropic effects.
Lance Oleny 04/24/18, Tuesday in Valentine #205-6
Title: "CoDrone: Consumer Friendly Programmable Drone"
Abstract: As automation and artificial intelligence continue to improve, the aviation industry is moving towards unmanned flight with drones. This talk will discuss my experience with programming a hand-held drone. Specifically, I have attempted to use speech-recognition software in python to send real-time basic flight commands to the drone. There will be code explanations in Python and C++. I will discuss my successes and limitation when working with the drone and will include a couple of demonstrations.
Esmeralda Quintana (SUNY Potsdam) 04/03/18, Tuesday in Valentine #205-6
Title: "Hack Potsdam"
Abstract: Esmeralda Quintana is a senior at SUNY Potsdam, pursing BA/MA in Mathematics and BA in Computer Science. She is the president for the SUNY Potadam ACM Chapter, and has been involved in hosting hackathons in Potsdam. She will speak about her experience of organizing hackathons in the local area, and also introduce upcoming events. For more information, please visit http://hackpotsdam.com/
Beau Yaremko 03/13/18, Tuesday in Valentine #205-6
Title: "Decision Tree Modeling for Heroes of the Storm Game Videogame"
Abstract: This talk will introduce the concept of decion tree modeling (also known as classification tree models) with applications to the Stat 113 first day survey and the video game, Heroes of the Storm. Heroes of the Storm is a competitive online video game that is growing in popularity. You pick a team of 5 heroes and battle against other players to advance through the map taking out enemy forts and towers until you finally make it to the other teams core. Once you destroy the other teams core you win the game. There are different maps with slightly different layouts that make players change up their strategies and team make-ups. Our goal is to use decision tree models to determine if there are certain strategies for each map that increase a team's probablilty of success.
Kira Murphy 02/27/18, Tuesday in Valentine #205-6
Title: "Hackathons: A Programming Event for Beginners and Experts Alike"
Abstract: What is a hackathon? According to Google, a hackathon is an event, typically lasting several days, in which a large number of people meet to engage in collaborative computer programming. Over President's Day Weekend, I attended Hack@CEWIT, a 48-hour event at Stony Brook University. I will be talking about what a hackathon is, how i ended up at this particular hackathon, what my team and I did, my overall experience, and upcoming hackathons near SLU.
Guinevere Gilman 02/13/18, Tuesday in Valentine #205-6
Title: "Database Design for Statistical Analyses"
Abstract: This talk will focus on the methods and concepts for building and designing a database to effectively store statistical data on hockey players' performance. It will include aspects of relational database design such as keys, constraints, and relations, and how a logical structure can be created to show the relationships between date while limiting redundancy. There will also be discussion of practical problems that I faced, such as populating the database with data from multiple sources.
Anna Izzo 01/30/18, Tuesday in Valentine #205-6
Title: "Sentiment, Analysis, Made Easy by R."
Abstract: In a world that is ever-evolving, people are no longer complaining about companies through their 1-800 numbers. Rather, if a company's products or services are disliked, consumers visit social media platfoms, such as Facebook, twitter, Instagram, etc.,and post their complaints in plain sight. Because of this, companies must show diligence in listening and responding to those who reach out in this manner.
Over the summer, Anna Izzo worked to understand and improve the efficiency and accuracy of determining sentiment through automation, insead of human power. From back end to front, she will share how she created a model in R to score over 15000 tweets in less than 20 seconds-now that's fast!
Malakia Takane 12/14 THRUSDAY in Valentine 202 - Please note the Day and Room change.
Title: Portfolio Utility Optimization
Abstract: Mathematical Finance is the field of mathematics that studies financial markets. My project focuses on developing an optimal investment strategy that seeks to build up wealth or pension payment stream while providing a balance between risk and reward. More precisely, the objective of this project is twofold: (1) to contribute to the practical application of utility optimization in building up wealth or retirement pension by utilizing an iterative procedure (a binomial tree model) and (2) to illustrate by way of case study the use of utility optimization to determine the percentage of capital in stock at each time period in the given optimal portfolio.
11/28/17 Allison Macri
Title: Fractals, Forgeries, and Forensics: Using Fractal Analysis to Identify Forged Handwriting
Abstract: This talk will introduce fractals and their relationship to forensic science. Examples of true and statistically self-similar fractals and their definitions will be given. This talk will also highlight the two important dimensions that classify fractals: the topological dimension and the Hausdorff-Besicovitch dimension, which is also known as fractal dimension. My research focuses on using fractal analysis to identify forged handwriting.
11/7/17 Nicole Williams
Title: Promoting Active Learning With Maple Trees
Abstract: My summer research focused on understanding the importance and the design of model eliciting activities. In the beginning of the summer I researched the design process and the important elements that are a part of model eliciting activities. Then I partnered with Nature Up North to look for a problem that was relatable and local to St. Lawrence University students that I could use for a model electing activity. In the end I created a model eliciting activity for college level classes, and have been able to implement it in statistics classes here at St. Lawrence University.
10/24/17 Yue Yang
Title: Counting Young Tableaux of a Given Shape
Abstract: My summer research focused on understanding the basics of Young diagrams and Young tableaux, and worked on the relationship between Young diagrams and Catalan numbers. Furthermore, this research discussed about how to introduce half-integers in Young tableaux, which becomes the main topic of my SYE.
10/10/17 Tiara Davis, of St. Lawrence is the Assistant Director of Career Services and Student Development
Some topics of discussion:How to get in touch with alumni, etc. regarding both full time employment and internship opportunities.
9/26/17 Nevaan Perera
Title: An Xbox Kinect Approach: Transforming Projector Screens into Interactive Smart-boards using the Xbox Kinect
Abstract: Smartboards add numerous benefits to the classroom experience such as easy writing and highlighting, surfing the web and controlling computer applications. These features significantly facilitate learning and teaching in the classroom; however, smartboards are with many shortcomings. Some of these shortcomings include cost, requirement of electronic pens and limitation of instructor participation. The current project intended to use the Xbox Kinect, which is a highly advanced depth sensor camera, to detect human body movements and execute tasks/applications on a projected screen. By leveraging such technology, we can transform regular projector screens into smart-boards, that are controlled just by the human body.
4/25/17 Last one for the semester
Soukaina Alami Idrissi
Abstract: Dailyhabit is an online web application which lets you keep track of your daily habits and long-term goals. I started this project during the web programming course with Prof. Lisa Torrey and turned it into a Senior Year project. I will be talking about how I acquired a domain and launched the website online, as well as the different tools and functionalities that I added to the web application
4/11/17 Madison Rusch and Haley Taylor
Title: Women in Cybersecurity Conference
Abstract: The Women in Cybersecurity Conference is a conference which aims to recruit, retain and advance women in the field of cybersecurity. With that goal in mind, they offer scholarships to undergraduate scholars so that students who might otherwise be unable to go to the conference have the opportunity to attend. This year, Madison Rusch and Haley Taylor, seniors in the Computer Science department, both received scholarships that allowed them to travel to Tucson, Arizona and participate in this year’s conference. They listened to keynote speakers, took part in technical presentations, and attended a Career Fair during their time at the conference. Come hear them talk about the process of applying for the scholarship, their time at the conference, and what they enjoyed most at this week’s Q-Club!
3/28/17 Greg Gardner and Gibson Drysdale
Title: Computing Shortest Paths Around SLU Campus
Abstract: Wouldn’t you like to know the shortest path from a location on campus to any and all other locations on campus? Is it possible? As partners, the topic of shortest paths hit home for us. Before starting this graph theory project, we figured that where we lived, 48 Park Street, was more on the outskirts of campus and therefore thought a shortest edges problem would be useful for us and many other people on campus. Using Dijkstra's algorithm and computer science, we have developed a way to turn the St. Lawrence campus map into a graph that tells us exactly this.
2/28/17 Kevin Frommer, Devin Robson, Andrew Volkmann
Title: NHL Ranking System
Abstract: This is a project that we did last year in our Graph Theory class. One of the types of graphs we learned about in the class was tournaments.
Specifically, how to take a tournament and rank teams based on wins and strength of their schedules. We choose to take this concepts that we learned and apply them to the NHL. We wrote an algorithm to create a ranking system for all NHL teams for both last year’s season and of all time. With the results that we found we were able to create a side by side comparisons with what we found to what the standings portrayed.
1/31/17 Gibson Drysdale
Title: Game-Making in Europe
Abstract: Not many off-campus programs offered by SLU give you the opportunity to get Computer Science credit. I was able to study abroad in Copenhagen, Denmark for four months in the spring of 2016, where I was enrolled in a Game Development program. The program teaches the basics of making video games and gives in-depth exposure into the European Video Game industry through company visits and other hands-on experiences. The program also counts for CS 300+ elective credit, which I needed if I wanted to graduate on time. I'll be covering the Denmark program, what it's like to live in Copenhagen, and the experiences I had during while taking this eye-opening and generally awesome course.
11/29/16 Elsa Fecke
Title: Comparing Statistical Methods for Classifying Brain Cancer
Abstract: The purpose of this project is to classify brain cancer subtypes using statistical methods and biological data. The statistical algorithms that are used to classify the cancer subtypes are naïve Bayes, K-Nearest Neighbors, Support Vector Machines and Random Forest. Two types of data are used in classification, including gene expression data and copy number data. The research question is: how can we obtain the best classification accuracy for prediction of brain cancer subtype?
11/15/16 Michelle Gould from St. Lawrence University Career Services will discuss the following topics
What they can do for you, what our majors have done in the past. How to get in touch with alumni, etc. regarding both full time employment and internship opportunities.
11/1/16 Sejla Palic will give a talk on her SLU Fellowship project she worked on with Jessica Chapman last summer
Title: " Pareto Front App: Optimizing Decision Making Process"
Abstract: Pareto Fronts are one way to simultaneously optimize multiple criteria. A choice is eliminated if it is worse than at least one other choice. This is done by considering “all” possible solutions/options and eliminating inferior ones. An option is eliminated from contention if it is worse than at least one other available option. Once a Pareto front is identified, only equally “good” choices are left to choose from. At that stage, the decision about which single option is “best” becomes subjective; however, this app provides some additional tools to guide you to a defensible solution. Even though functions for finding Pareto fronts are written in R language and are available to the public, an average person wouldn’t be able to use them efficiently. The purpose of our project was to make a web app that would allow users without R knowledge to utilize Pareto Fronts functions on large data sets.
10/18/16 Holley James-Grisham will speak on his research
Title: " Using Stylometry to Classify Movie Script Genres with Each Other"
Abstract: Stylometry is the statistical breakdown of deviances in literary style among writers or genres (Burrows, 267). Stylometry can be used in situations where people come across books that they like, and search out the author in hopes that the author has written similar books. After days of searching they might realize that the author of the book is unintentionally unidentified. The book is from ancient Greece and the author’s page got lost over time. However, a statistical algorithm that uses the words from the book to correlate it to one specific author could be used to discover who wrote the book. The theory of stylometry will be tested using movie scripts from the web. The main question is whether or not different stylometry methods can be used to distinguish between different genres. Once the basic steps to filtering through the flooded data is successfully complete using computer coding software in R, the data will then be ready for analysis using a variety of stylometry techniques to represent each data set. Finally, a variety of numerical representations from the stylometry techniques will be used to create cluster plots to see the clusters of similar data sets. This research will allow writers to group different pieces of work to different time periods, authors, and possibly genres. The likeliness that research is successful will be determined by whether or not there is an accurate correlation between different data sets.
10/4/16 speaker Brian Coakley (CFO @ North Country Savings Bank),
Mr. Coakley will talk about his career path from a computer science major to becoming CFO at a bank.
9/20 16 Yuxi Zhang (Abstract Fellowship 2016)
Title: " User Interface and Usability Analysis for Importing Data in Statistics & Data Analysis Tools"
Abstract: As statistical data analysis software tools are used intensively and extensively during research and decision-making in industries, software engineers also never stop seeking more user-friendly interface designs to prevent cases where users fail to import external data into those analysis tools effectively. The format of the data and the size can make data import challenging.
My fellowship project focused on improving the data import process of a web-based tool “StatKey” with a more user-friendly interface. StatKey is used extensively by students taking Applied Statistics courses here at SLU and many other users worldwide.
9/6/2016 Taylor Pellerin (Abstract Fellowship 2016)
Title: "Play Select Like a Champion: Run/Pass Decision Making in FBS College Football"
Abstract: Before every single offensive play in American football, the coach has a choice to make: which set of instructions to give the team. Each play call is a decision that is tactically made. Depending on any number of conditions, a coach will have the team run or pass the ball, with the goal of advancing the field and ultimately scoring. After constraining and cleaning up a very large data set, I built a set of linear models which look at the success of certain factors in predicting efficient play calling. In this session, I will talk about the ways in which the data set was manipulated, the process of developing the models, as well as a change in run/pass decision-making that the results of the models point to.