Marco Gaboardi (Boston University)

Data analysts aim at guaranteeing that the result of a data analysis run on sample data does not differ too much from the result one would achieve by running the analysis over the entire population. To achieve this goal, they have developed several techniques to control the generalization errors of their data analyses. In this talk, I will discuss how programming language techniques can help data analysts to design adaptive data analyses...

Yingyu Liang

An important characteristic of neural networks is their ability to learn representations of the input data with effective features for prediction, which is believed to be a key factor to their superior empirical performance. To better understand the source and benefit of feature learning in neural networks, we consider learning problems motivated by practical data, where the labels are determined by a set of class relevant patterns and the inputs are generated...

We introduce quivers, path algebras and their representations. Then, in the case when our ground field is algebraically closed, we discuss a particular Morita invariant of path algebras arising from finite quivers, the Ext quiver of the category. Through examples we see how to compute the Ext quiver using quiver representations and techniques from linear algebra. We aim to keep the talk accessible to undergraduate and graduate students alike.

Ryan Bianconi UI Mathematics PhD...

From a summer REU at the University of Minnesota, we constructed a solvable lattice model for the dual weak symmetric Grothendieck polynomials in hopes of using such a model to prove related properties of these polynomials, including Cauchy identities and branching rules. We also considered a similar lattice model construction for the weak symmetric Grothendieck polynomials in hopes of proving a Cauchy identity, concluding with a negative result. Moreover, we expand on previous work by...

Digital signatures are an important feature in any encryption/decryption scheme, as it provides a message with integrity, authenticity, and nonrepudiation. The problem occurs when long messages are being exchanged and signatures that are just as long need to be verified. By using hash functions, a ”fingerprint” of the message can be used instead of the message itself for verification, making the process computationally inexpensive. If we consider a single iteration of a general hashing...