Spring 2015 Events
Wednesday, April 22, 12:15-1:00, 416 Clapp--Talk Back
The last Math/Stat Club meeting of the year is our annual “Talk Back” -- an opportunity for you to give the department advice about opportunities and events outside the classroom. What are your reactions to opportunities and events in 2014-2015? What advice/suggestions do you have for 2015-2016? Come share your ideas and join the conversation.
Wednesday, April 15, 12:15-1:00 pm, 416 Clapp--Speaker: Rudo Mudzi '14, Data Scientist, MassMutual Financial Group
Disability Income Insurance Morbidity Modeling at MassMutual
In the recent years, rich data has increasingly played a pivotal role in how both the private and public sectors make their decisions. MassMutual Financial Group takes advantage of its rich data by building predictive models to make better-informed decisions in running its business. In this talk we will look at survival analysis as a method of predicting incidence rate, as well as other claim experiences for incoming disability income insurance cases at MassMutual
Wednesday, April 8, 12:15-1:00 pm, 416 Clapp--Speaker: Raluca Ursu '10, PhD Student, University of Chicago
The Power of a Ranking: Quantifying the Effects of Rankings on Online Consumer Search and Choice
In this talk I will tell you about my journey from a Math and Economics major at Mount Holyoke College to a PhD student in Economics at the University of Chicago. For my thesis, I study how consumers’ choices are influenced by the order in which products are presented to them. I focus on an experiment done at Expedia where consumers saw hotels ranked in a random order instead of Expedia’s proprietary ranking. This experiment is exciting because it allows me to identify the causal effect of a ranking on purchases, which was previously not possible, as well as study how consumers’ online search is influenced by the ranking they observe. As we advance in an era of information abundance that has to be summarized in order to be useful, understanding how rankings affect choices is paramount. I will focus on how I am using math to study consumer behavior, as well as venture back to my MOHO days and discuss relevant coursework and preparation for those of you who are curious about graduate school.
Wednesday, April 1, 12:15-1:00 pm, 416 Clapp--Speaker: Luong Nguyen '12, Graduate Student, Joint CMU - UPitt PhD program in Computational Biology
Segmenting salient objects: How does statistics bring African safari closer to cancer tumor?
Segmentation is a difficult problem in computer vision. Numerous methods have been developed over the last few decades to tackle this problem. However, there are only a limited number of available methods specially customized for tissue images, in particular ones of biopsied tumors. The biopsied tumors are crucial to treatment decisions for cancer patients. A key component of biopsied tissue analysis is to identify diagnostically relevant architectural elements in a tissue. Despite the significance of this task, most of the currently available segmentation methods are not sufficiently robust and scalable due to their ulitization of highly engineered features and requirements of extensive training data. Recent developments in segmentation of natural images (e.g. African safari) show that statistical representation has the potential to overcome the aforementioned challenges. I will demonstrate a new segmentation approach inspired by state-of-the-art computer vision methods using images of breast tissues.
Wednesday, March 25, 12:15-1:15 pm, 305 Kendade--Math/Stat and CS Joint Talk
Speaker: Marc Maier, MassMutual Data Science
Friday, March 13, 9:26:53 AM, 416 Clapp-- Pi Day Celebration
Hello Pi(e) Lovers!
Join us for a brief, but fun, (Pre)Pi Day Celebration in 416 Clapp on Friday in 416 Clapp at 9:26:53 AM. There will be pie for all, and a surprise treat for the first arrivals, until supplies run out.
Our celebration is taking place exactly 24 hours before the Epic Pi Day of the Millennium --- this year we will be celebrating Pi to 9 decimal places: 3.141592653 --- in other words, March 14, 2015 at 9:26:53 AM. This won't happen again until the next millennium.
Come a few minutes beforehand for the countdown and give a shout out to Pi --- and join the pi(e) circle!
Wednesday, March 11, 12:15-1:00 pm, 416 Clapp--Speaker: Amelia McNamara, UCLA
Considering Mathematics or Statistics for your major or minor? Come join us for pizza and conversation!! Current majors and minors are invited to come share their experiences with you. Hope to see you there!
Wednesday, February 18, 12:15-1:00 pm, 416 Clapp--Speakers: Dara Zirlin and Samantha Vanschalkwyk
Wednesday, February 11, 12:15-1:00 pm, 416 Clapp--Speaker: Eric Reed, UMass Graduate Student
A brief discussion of interdisciplinary studies combining mathematical principles with life sciences
Geometry and the complexity of computation
Wednesday, January 28, 12:15-1:00 pm, 416 Clapp--Speaker: Grace Barkhuff '16, Research Through the Budapest Semesters in Mathematics REU Class
Using Randomization to Understand Neural Networks
The brain is comprised of millions of neurons, connected in particular patterns in order to send electrical impulses and messages between one another. These neurons and connections are often visualized as graphs comprised of points and lines, called vertices and edges. On the graph, vertices represent individual neurons and edges represent the connections between them.
While sleeping or resting, the hippocampus of the brain sends electrical impulses in a Sharp Wave Ripple (SWR) pattern which helps the brain form memories. Given a specific set of data which represents a neural network of the hippocampus of a rat brain, my research group used computer simulations to understand which particular aspects of the structure of the neural network were vital to have a functioning hippocampus, as determined by the presence of SWR. This was done using randomization processes on an adjacency matrix of the neural network in order to rule out one at a time which aspects of the network are not necessary to produce SWR.
Fall 2014 Events
Wednesday, December 3, 12:15-1:00 pm, 416 Clapp--Where in the world can math/stat majors go?
Please join us for the final Math/Stat Club meeting of the semester. As this semester comes to a close, a lot of us are already thinking about next semester and beyond. At this meeting we will talk about opportunities for study abroad for majors in mathematics and statistics and any special planning necessary. Students who have returned from study abroad experiences will be on hand to talk about their experiences and help answer questions.
Wednesday, November 19, 12:15-1:00 pm, 416 Clapp--Speaker: Chaitra Gopalappa, Mechanical and Industrial Engineering, UMass
Systems Simulation Modeling for Public Health Applications
Infectious diseases such as HIV, TB, malaria, and Ebola disproportionately affect certain populations. Though the how-tos of prevention are known, their implementations are challenging because of the complicated interactions between disease, socio-economics, and behavior. A holistic approach to disease prevention often includes interventions at the individual-level, such as testing programs to diagnose and treat infected persons, interventions at the population-level, such as awareness programs, and interventions at the systemic-level, such as timely updates to guidelines and policies. Interventions may also include continuing care at various stages of disease progression to minimize disease burden. Several of these interventions differ by population groups based on need.
Given these complicated interactions, techniques that can model the dynamics of these interactions, such as systems simulation models, are used to evaluate intervention scenarios so the limited resources can be allocated efficiently. Analyses of short- and long- term impacts and costs of alternative scenarios of combination intervention programs could inform the development of strategies for disease prevention. In this talk I will discuss two types of systems simulation modeling techniques, compartmental and agent-based, for application to public health decision making.
Have you considered an actuarial career but you are not too sure what actuaries do in practice? With a 26% projected growth in employment from 2012-2022*, the possibility of undertaking this profession is worth exploring. Come learn more about this exciting profession and its challenges. In this talk, I will present an overview of the typical roles of actuaries, who their employers are, what the examination process is and finally, a brief description of my personal experience as a retirement plan actuary.
* Bureau of Labor Statistics, U.S. Department of Labor, Occupational Outlook Handbook, 2014-15 Edition, Actuaries
Euclidean proofs of the infinitude of primes in arithmetic progressions
Wednesday, October 29, 12:15-1:00 pm, 416 Clapp--Spring Classes!!
Please join us to hear about 300-level classes offered here on campus and around the 5 colleges this spring.
Wednesday, October 15, 12:15-1:00 pm, 416 Clapp--Speaker: Samrachana Adhikari, MHC '11
Statistical Analysis of Real-World Networks
Network data representing complex phenomena in fields such as biology, physics, economics, finance and education is abundant today. Social online networks (e.g. Facebook, Twitter), brain networks, gene networks, financial networks, friendship networks of teachers in schools, etc. are some of the examples of such real-world networks. Researchers hope to understand underlying complex phenomena that form such networks in these fields by analyzing observed network data. A network consists of fixed nodes (e.g. social online users) and dependent edges (e.g. friendship relations). Many real-world networks are noisy and very large. Modeling stochasticity in such highly dependent complex structures poses challenges in existing statistical approach and is computationally harder. Most of the recent works in statistics to analyze networks has been on this front. In this talk, I will delve into the need to build consistent and interpretable statistical models to analyze real-world networks in a systematic and computationally feasible manner. I will motivate the talk by giving examples of networks from different domains, as well as the associated questions researchers in these domains attempt to answer. Next, I will give a general overview of the existing methods for network analysis. In particular, I will discuss ongoing work in the Hierachical Network Methods (HNM) research group in the Statistics Department at Carnegie Mellon University.
Wednesday, October 8, 12:15-1:00 pm, 416 Clapp--Reports on summer activities from Math/Stat students, Part III
Wondering what kind of opportunities are available for students interested in mathematics and statistics? Please join us to hear students Adelyn Yeoh, Caiyun Zhu, Anne Waldo, Dara Zirlin, Maria Ferreras, and Stephanie Stark share information about their summer experiences.
Adelyn, Caiyun and Anne worked together in a number theory research group on campus. Stephanie also stayed on campus and worked on problems in rigidity theory related to computer-aided design. Dara went to the REU at Lafayette College and Maria was an intern at Bank of America Merrill Lynch in Equity Capital Markets.
Wednesday, October 1, 12:15-1:00 pm, 416 Clapp--Reports on summer activities from Math/Stat students, Part II
Wondering what kind of opportunities are available for students interested in mathematics and statistics? Please join us to hear students Taylor Anderson, Mariah Mullins and Stephanie Stark share information about their summer experiences.
Taylor participated in the Complex and Dynamical Systems REU at the University of Minnesota. Mariah did research with a professor of statistics at UMass who works with network methods used by the CDC and UNAIDS. Stephanie stayed on campus and worked on problems in rigidity theory to computer-aided design.
Wed., September 24, 12:15-1:00 pm, 416 Clapp -- Reports on summer activities from Math/Stat students, Part I
Wondering what kind of opportunities are available for students interested in mathematics and statistics? Please join us to hear students Sam VanSchalkwyk, Amanda Benson, and Maria Sellanes share information about their summer experiences.
Sam participated in the MSRI-UP research program. Amanda taught math and art classes to children in Nicaragua. Maria worked as a camp counselor at MathPath.
Wed., September 17, 12:15-1:00 pm, 416 Clapp -- The Math/Stat Book Club presents: The Man Who Knew Infinity
Please join us for our first Math/Stat Club meeting of the year. This is an opportunity to meet faculty and other students interested in mathematics and statistics. You will hear about faculty research interests and many opportunities for students to meet each other and get involved in our department.
The last Math/Stat Club meeting of the year is our annual “Talk Back” -- an opportunity for you to give the department advice about opportunities and events outside the classroom. What are your reactions to opportunities and events in 2013-2014? What advice/suggestions do you have for 2014-2015? Come share your ideas and join the conversation.
Wed., March 26, 12:15-1:00 pm, 416 Clapp -- 300-level courses
Wondering what to take next fall? Come talk with Math and Stat faculty members about the 300-level courses they will be teaching.
Wed., March 5, 12:15-1:00 pm., 416 Clapp -- Speakers: Math Modeling Competition Participants
The speakers at this week's Math/Stat Club will be student participants in this year's International Mathematical Modeling Competition. They will discuss this year's contest problems and their solutions, and their experiences with mathematical modeling over the past year. If you are interested in learning more about mathematical modeling and the opportunities for students to get involved, come hear them speak and ask them questions.
Wed., February 26, 12:15-1:00 pm, 416 Clapp -- Speaker: Professor Jessica Sidman, MHC
Perspective in Two Dimensions for Computer Graphics
Holding in mind a 2D projection of a 3D scene is easy; holding in mind the 3D scene is hard. According to recent research in perception the human brain deals more easily with 2D representations than with 3D ones. When perspective was re-invented in the Renaissance artists quickly discovered how to construct accurate perspective without having to work in 3D. Doing so they were able easily to combine perspective with composition, which is dominantly 2D. From this point of view, modern computer graphics seems gratuitously difficult in forcing its users to develop full 3D models.
The seminal construction in Renaissance artists' perspective is the tiled floor, the many constructions of which I will explain in detail. Having reduced it to 2D perspective geometry, I will show how other useful constructions can be similarly reduced, including the double projection required for accurate shadows. Artists did these computations geometrically, drawing lines to find their intersections. Using my formalism to reduce geometry to algebra, computers can reproduce the 2D perspective calculations used by artists.
Wed., February 12: No scheduled talk
Friday, February 7, 4:00 pm, 416 Clapp -- Speaker: Brad Westgate
Introduction to Statistical Data Mining
Data mining refers to several types of problems, all involving extracting information from a large dataset. We will review different branches of data mining, including classification, clustering, and dimension reduction, and consider example applications where these problems arise. We will focus in particular on classification, which involves predicting an unknown outcome from several other observed variables. We will discuss and compare classification methods, including Naive Bayes and logistic regression.
Friday, February 7, 12:15-1:00 pm, 416 Clapp -- Speaker: Andrew Bray
Tectonics and Tessellations: residuals for spatial point processes
Earthquakes are one of many natural phenomena that can be effectively modeled as a spatial point process. However, challenges abound. In this talk we will outline the process by which earthquake models are formulated and discuss the difficulties that arise in assessing how well they perform. We will also investigate how spatial tessellations can lend substantial power to the modeling process.
Friday, January 31, 12:15-1:00 pm, 416 Clapp -- Speaker Kelly McConville
If a tree grows in a forest, will a statistician count it? An introduction to survey sampling
Surveys are fairly common in our society and allow one to make inferences about a population of interest based on sample. The sample data, collected using a complex sampling design, can often be supplemented with additional data. In this talk I will discuss how to estimate a population quantity in the presence of both survey and additional data. In particular, I will present an estimator I developed -- the LASSO Survey Regression Estimator -- and I will address how this estimator compares to other commonly used survey estimators when estimating tree canopy cover for a region in Utah.
Monday, January 27, 12:15-1:00 pm, 416 Clapp -- Speaker: Andrea Foulkes
Statistical deception and the era of "big data"
The familiar phrase "Lies, damn lies and statistics" takes on new meaning in the era of "big data". In this Math/Stat club talk, we will begin by exploring various ways in which statistics and statistical graphics have been used to persuade and, at times, deceive us. Focus will be on the thematic use of several important statistical concepts, ranging from misrepresentation of a mean to taking advantage of the presence of selection bias. We will then consider how big data applications present an entirely new opportunity to use these old tricks, as well as a few new ones. Students (and faculty) will learn to have a more discerning eye, as well as learn about why understanding and applying elementary statistical concepts is so important in the big data arena.
Fall 2013 Events
Wed., December 4, 12:15-1:00 pm, 416 Clapp
End-of-Term party: Come celebrate the last Math/Stat Club meeting of the semester
Wed., November 20: Event Cancelled
Wed., November 13, 12:15-1:00 pm, 407 Clapp (note location)
"Enabling Secure Computing through Fully Homomorphic Encryption"
Come join us for an very interesting cross-disciplinary talk by Kurt Rohloff of BBN Technologies.
One of the major breakthroughs of theoretical computer science in the 20th century was the formalization and development of mathematically secure encryption technologies. These breakthroughs have enabled people to securely share information over the Internet without prior interactions. In this talk we'll provide an introduction to prevalent encryption technologies such as one-time pads, AES-based symmetric key encryption and RSA-based public key encryption technologies. We'll particularly focus on practical trade-offs associated with building and using these technologies. Building upon this prior work, we will also discuss the first major breakthrough of computer science in the 21st century - the demonstration of public-key Fully Homomorphic Encryption (FHE). FHE allows sensitive data to be encrypted such that arbitrary programs can be securely run over the encrypted data where the output, when decrypted, is equivalent to the result of running the original algorithm on the unencrypted data. FHE is an area of high ongoing interest from both theoretical and practical perspectives.
Wed., November 6, 12:15-1:00 pm, 416 Clapp
300-Level Course Meeting
Faculty members who are teaching 300-level courses in the spring semester will discuss their offerings. It will be a chance to think over your math majors and minors and ask any questions you have on those.
Wed., October 23, 12:15-1:00 pm, Clapp 407 (Note change of room)
Rigidity Theory for Robotics, Drug Design and CAD
Jessica Sidman and Audrey St. John
When designing a bridge, how can we minimize the amount of building materials while maintaining stability? Are there computational tools that can help predict protein flexibility, a key component in drug design? Rigidity theory seeks to answer these questions by studying structural properties from the mathematical and computational perspectives. In this talk, we will discuss the fundamental questions considered by rigidity theory and its applications, including those in robotics, structural biology and Computer Aided Design for mechanical engineers. We will provide an overview of the mathematical tools used in analysis and discuss opportunities for getting involved in ongoing research projects.
Wed., October 9, 12:15-1:00 pm, 416 Clapp
Presentations on Summer Work by Fellow Mathematics and Statistics Students Part II
Wed, October 2, 12:15-1:00 pm, 416 Clapp
Presentations on Summer Work by Fellow Mathematics and Statistics Students Part I
Wed, September 25, 12:15-1:00 pm, 416 Clapp
A Look at the World of Finance, Zheyuan Hu, BlackRock Financial Management in NYC
The Problem Solving Seminar is an off-shoot of the Math/Stat club. Students who intend to take the Putnam along with others interested in improving their problem-solving abilities meet weekly on Friday for lunch noon to 1:00 in Clapp 416 (in the fall semester) with a faculty member to work on problems. The seminar is not for credit, it is an extra-curricular activity. Read this article about our problem solving group Some Truly Impressive Figures, May 2, 2011.
This year the club will hold its first meeting on Friday, September 28, problems get sent out on Sundays for group members to look over. We alternate weeks between solving problems one week and practicing our presentations the next week. On the first Saturday in December members from the Problem Solving Seminar take the Putnam exam.
- Mathematical modeling group (Advisor : Dylan Shepardson)
Are you interested in learning to use mathematical methods to solve real world problems? The Mathematical Modeling Club meets weekly to discuss modeling strategies and work together on problems. Members of the Mathematical Modeling Club will elect a 3-person team to represent Mount Holyoke College in an international mathematical modeling contest that takes place in February. You do not need to be a mathematics or statistics major. Being able to think creatively about how to solve real problems is the most important thing, and the group benefits from having a variety of perspectives. People from other disciplines are welcome!
- GRE group (Advisor: Giuliana Davidoff)
If you are planning to take the GRE Subject Test in Mathematics this fall (October or November), you should be working problems in as many sample tests as you can find. You should also be giving yourself timed practice tests, since there is no substitute for that experience. If you have questions about any aspect of this exam, whether about particular problems or particular topics, be sure to get those answered.
You're more than welcome to join the Mount Holyoke Actuarial Club if you're interested in becoming an actuary and want to learn a great deal about the insurance industry!
The Actuarial Club at Mount Holyoke College aims to help its members become strong actuarial candidates, to provide resources necessary for them to prepare for an actuarial career, and also to build a professional network with our wonderful alumnae actuaries.