alexander iannantuono

Myself, circa 2016

http303 -> 127.0.0.1

Kelowna, BC

Canada, V1V 000

I am a MSc student in mathematics at the University of British Columbia Okanagan Campus. I am working under the supervision of Yves Lucet. I graduated virtually from McGill University with a BSc in mathematics and computer science.

Research

Current: nothing for now! just taking neat classes.

In the near future: optimization for road design applications

Timeline

2020

Worked on a neat class (MATH562 at UBCO with Warren Hare) project which involved investigating surrogate problems in (black box) optimization.
- I wrote a small document to gather my thoughts (as well as data and figures) in preparing for a presentation.
- Coding in MATLAB + Python was involved, code is available on GitHub
Worked on a project that aims to provide a measure of confidence that a deep neural network has given some classification task
- Pre-print on arXiv
- Recently chosen for a spotlight presentation at the ICML 2020 Uncertainty & Robustness in Deep Learning (UDL) workshop

2019

Worked with Aram-Alexandre Pooladian on a project involving Deep Learning (DL) perturbation attacks using a Langevin dynamics based approach (preprint)
Worked on an algorithm that became the No-collision transportation maps paper
- It fares well empirically for approximating the optimal transport solution in \(\mathbb{R}^n\), \(n > 1\) using \(\ell_1^2, \ell_2^2, \ell_\infty^2, \ell_2\) as measures
- A poster (pdf) was presented the NeurIPS 2019 OTML workshop in December 2019)
- Pre-print on arXiv, code on GitHub

Interests

Optimization
Optimal transport
Numerical methods
Machine learning
Distributed and parallel computing

I hope to specify further as stronger interests develop.

Small Talk Topics

Some hobbies of mine that do not pertain to the above

Bicycles: cycling, maintenance, etc.
Collecting (vinyl) records, high-fidelity sound
Almost anything digital: repair, building, design, etc.
Photography (sometimes with artistic distortion), graphic design and typography
Tinkering in the vain of N-O-D-E’s projects

Nationality: Canadian

Languages (spoken + written):

English
French
~Italian (looking to improve)

Colleagues & Friends

news

Oct 23, 2020	I will be giving a talk on the No-collision transportation maps work at the COCANA on ~~January 14~~ March 4, 2021 via Zoom. The slides for the talk that I gave can be found here.
Jul 5, 2020	Work on using Out-of-Distribution (OOD) data to calibrate deep neural networks has been accepted to the UDL workshop!

selected publications

JSC

No-Collision Transportation Maps

Nurbekyan, L., Iannantuono, A., and Oberman, A. M.

2020

Abs HTML PDF

Transportation maps between probability measures are critical objects in numerous areas of mathematics and applications such as PDE, fluid mechanics, geometry, machine learning, computer science, and economics. Given a pair of source and target measures, one searches for a map that has suitable properties and transports the source measure to the target one. Here, we study maps that possess the no-collision property; that is, particles simultaneously traveling from sources to targets in a unit time with uniform velocities do not collide. These maps are particularly relevant for applications in swarm control problems. We characterize these no-collision maps in terms of half-space preserving property and establish a direct connection between these maps and binary-space-partitioning (BSP) tree structures. Based on this characterization, we provide explicit BSP algorithms, of cost O(n log n), to construct no-collision maps. Moreover, interpreting these maps as approximations of optimal transportation maps, we find that they succeed in computing nearly optimal maps for q-Wasserstein metric (q = 1, 2). In some cases, our maps yield costs that are just a few percent off from being optimal.