Universality in numerical computations with random data. Case studies.

Percy Deift
New York University
October 13, 2016
This is joint work with Govind Menon, Sheehan Olver and Thomas Trogdon. The speaker will present evidence for universality in numerical computations with random data. Given a (possibly stochastic) numerical algorithm with random input data, the time (or number of iterations) to convergence (within a given tolerance) is a random variable, called the halting time.

The Left Side of History: World War II and Re-emergent Nationalisms in Contemporary Eastern Europe

Kristen Ghodsee
Professor and Director of Gender, Sexuality, and Women’s Studies at Bowdoin College, former Member (2006–07) in the School of Social Science, and President of the Association of Members of the Institute for Advanced Study
October 14, 2016
Kristen Ghodsee will investigate the contemporary European memory projects about World War II and the Cold War. Since the global financial crisis in 2008, countries once locked behind the Iron Curtain have increasingly drifted to the far right, vilifying their state socialist pasts to exonerate nationalist heroes once condemned for their collaboration with Nazi Germany. Politicians and scholars strategically deploy historical knowledge as a tool to quash growing domestic opposition to the economic upheavals and insecurities of the post-socialist era.

Nessun Dorma: From Night Stories to a History of the Night in the Greek World

Angelos Chaniotis
Professor in the School of Historical Studies
October 14, 2016
Angelos Chaniotis will discuss the night, whose definition as the period between sunset and sunrise is consistent and unalterable, regardless of culture and time. However, the perception of the night and its economic, social and cultural roles are subject to change. Which parameters determine these changes? What can we learn by studying them about the specific character of a culture? Why do people experience the night in different ways in different historical periods and how did this affect their lives?

Matrix invariants and algebraic complexity theory

Harm Derksen
University of Michigan
October 17, 2016
The determinant of an $n\times n$ matrix is an invariant polynomial of degree $n$ that is invariant under left and right multiplication with matrices in ${\rm SL}_n$. It generates in the sense that every other invariant polynomial is a polynomial expression in the determinant. In this talk we consider the simultaneous left and right action of ${\rm SL}_n$ on $m$-tuples of $n\times n$ matrices. I will explain a joint result with Visu Makam that shows that invariants of degree $\leq n^6$ are sufficient to generate all polynomial invariants.