Thursday, November 14, 2013
Bonnie Ray, IBM
Business Analytics Research at IBM
As one of the largest IT companies in the world, IBM has been at the forefront of systems, software, and physical and mathematical sciences research for many decades. However, only in the last ten years has it begun to systematically apply advanced statistics and optimization capabilities to support its internal business decision making. In this talk, I’ll provide examples of how the Business Analytics and Math Sciences organization within IBM Research has partnered with other IBM divisions to enable data-driven business transformation. For one example, that of predicting performance of new business initiatives, I’ll discuss two different statistical approaches investigated, Regression Trees and Nearest Neighbors with Metric Learning, and discuss their performance from both a technical and a business perspective. I’ll also highlight some open problems with potential for future research.
4:00 pm in Seeley Mudd 206 with refreshments in Seeley Mudd 208 at 3:30.
Tuesday, November 19, 2013
Summer REU Information Session
Interested in Summer Research Programs in Mathematics and Statistics?
There are many opportunities to do research in all areas of the mathematical sciences during the summer. Many of these programs are called REUs (Research Experiences for Undergraduates) and they happen at colleges and universities all over the country. Most give you accommodation and also pay a stipend for living expenses, so it's a pretty good deal, and lots of fun!
There will be a chance to hear from Amherst students who participated in on-campus and off-campus programs, as well as to ask questions about a variety of programs.
4:00pm in Seeley Mudd 205 (cookies provided).
If you cannot come to the session or want more information now, you can look at the Math Department web page at:
We'll be updating this with new opportunities as we hear about them.
Thursday, November 21, 2013
Haynes Miller, MIT
Kervaire Invariant One
In 1938 Lev Pontryagin and Hans Freudenthal published differing computations of pi4(S2). The matter rested till 1950, when George Whitehead resolved the matter (in Freudenthal's favor). The dispute involved the value of a certain invariant which we see today as a case of the "Kervaire invariant," and is an early illustration of the subtlety of this invariant. The work of Kervaire and Milnor in the early 1960's reduced fundamental questions about high dimensional manifolds to homotopy theory, and their answer involved the Kervaire invariant. As other geometrically inspired questions in homotopy theory fell to the tools of algebraic topology, the Kervaire invariant question emerged as the knottiest one of all. So there was celebration in April, 2009, when Hill, Hopkins, and Ravenel anounced a resolution of this question (except in one dimension). I will discuss this problem, its history, and something of what went into the proof (which is much of the development of homotopy theory over the intervening half century).
4:30 pm Seeley Mudd 206 with refreshments in Seeley Mudd 208 at 4:00.
This is the annual Connecticut Valley Mathematics Collequium talk and will be followed by a dinner, reservations required. To RSVP email firstname.lastname@example.org, indicating your choice of chicken or vegetarian entree.
Talk on Monday, March 24, 2014 by Lisa Dierker (Wesleyan University)
Taking a Passion for Statistics to the Classroom, into the MOOC World and Back Again
This presentation will describe an NSF-funded, multi-disciplinary, project-based model for teaching introductory statistics that directly and creatively tackle many of the most significant challenges currently faced by introductory statistics instructors and students. The curriculum is aimed at taking advantage of students' natural curiosity and providing a common language for approaching questions across numerous scientific disciplines. Core features of this curriculum include providing opportunities for students to flexibly apply their knowledge, the use of real world data, computing as a window to core statistical concepts, and supporting students with varying levels of preparation. I will also include information on the course as a MOOC and how experiences with that delivery format is now shaping our thinking about the classroom experience at Wesleyan and our partner institutions.
4:30 pm in Seeley Mudd 206 with refreshments in Seeley Mudd 208 at 4.
DATA FEST: Weekend Event March 28-30
Last year, nearly 100 students participated in the 2013 DataFest at UCLA, and many others took part at Duke University. This year, DataFest is coming to the Pioneer Valley in the form of the Five College DataFest, to be hosted at UMass-Amherst. Undergraduates from throughout the Five Colleges (Amherst, Hampshire, Mount Holyoke, Smith and UMass) will have the opportunity to compete in teams of up to five students, to address a real-world problem supported by an interesting data set. The Five College DataFest will take place over the weekend of March 28th at UMass. Will you rise to the challenge of making sense of the data? Please let us know if you are interested in meeting informally in February to get introduced to some technologies for visualization and data management which might be useful. More info can be found at: http://www.science.smith.edu/departments/math/datafest/
New and Immediate Change to Stat 111 for Fall 2014
Due to demand, Stat 135 will be offered BOTH semesters next year, but this has required converting Stat 111 Section 04 to Stat 135 for the Fall. Please check the section you are attempting to pre-register for and adjust your selection as necessary. Stat 135 has a calculus pre-requisite, while Stat 111 does not.
To reiterate: Stat 111 Section 04 in the Fall has been converted to Stat 135. This will be official in the course scheduler in the afternoon on Monday, April 7th. Please plan accordingly!
Amherst mathematics and statistics majors participate in sports analytics.
Nine Amherst students participated in the DataFest event held March 28-30 hosted at UMass Amherst on two teams: Speed Data-ers and Dataminerz. Dataminerz won one of the three awards presented for their work detecting errors in the data. Congratulations to all the students who participated! For more information go to http://magazine.amstat.org/blog/2014/06/01/datafest/
Research on Reproducibility in Statistics Curriculum Published
Prof. Nicholas Horton and collaborators (some local, some not) recently published work on incorporating reproducibility in the statistics curriculum. Using R Markdown, students are able to streamline their analysis and presentation of results steps, rather than taking results from a software package and then formatting a document in Word to convey their results. Seeing the entire workflow from start to finish can make it easier to detect mistakes and avoid issues with selective reporting of results, as well as help students understand the analysis process. To read more about this research, visit: http://today.duke.edu/2014/02/reproducibility
Pi Day Celebration
The recently re-named Department of Mathematics and Statistics celebrated Pi Day with a social gathering with pizza (pie) and (dessert) pies. For pictures of the event, courtesy of Anne Torrey, look here.
Information Session for Prospective Majors in Mathematics and/or Statistics
When: Monday, March 31st at 7:30 pm
Where: SM 207
Who: Professors R. Benedetto and Wagaman, several current majors, and YOU, our prospective majors
What: Brief overview of mathematics and statistics majors - requirements for the new statistics major, changes to the mathematics major, how honors works, etc. Then time for your questions - whether those are formal questions about requirements or informal questions about courses. Finally, some social time and refreshments for prospective majors to mix with some current majors and faculty.