Alejandro Flores V.

PhD candidate in Computer Science, UMD

I’m a PhD candidate of the Department of Computer Science at the University of Maryland, College Park, advised by Prof. David Mount. My research interests lie generally in Theoretical Computer Science, including Computational Geometry, Approximation Algorithms, and Data Structures. I’m always looking for research opportunities, just contact me!

My current research revolves around the goal of achieving more efficient nearest-neighbor classification, showing results that are practical for large-scale applications, while also having provable theoretical guarantees. This work focuses on two main subjects. First, on the development and analysis of algorithms to reduce the training set used for classification. These reduced subsets can be either heuristically obtained [CCCG’19, CCCG’20] or \(\varepsilon\)-coresets [ESA’20]. Additionally, I’ve been working on proposing new data structures specially tailored for answering \(\varepsilon\)-approximate nearest-neighbor classification queries [ESA’21], achieving the lowest complexity of any approach so far.

I completed my undegraduate studies at Universidad Simón Bolívar, Venezuela, earning a BEng. in Computer Science, Cum Laude. During those years, I had the pleasure of working with great people. First, with Prof. María-Esther Vidal, on topics related to Semantic Web and Graph Databases (see publications). Additionally, I was advised by Prof. Emely Arráiz towards my undegraduate thesis, titled Bio-Inspired Metaheuristics for Instance Selection, which was awarded an outstanding honorable mention by the committee.

Born and raised in Venezuela :venezuela:. My country lives under a dictatorship that has forced over 5M people to flee. Venezuela deserves to be free, and we will never stop fighting. During my free time, I enjoy photography (specially macro photography), biking, hiking, dancing, and cooking.

Condensation Heuristics

See CCCG'19, CCCG'20

$$ \varepsilon $$-Coresets for the NN Rule

See ESA'20

Chromatic AVD

See ESA'21

Education

Ph.D. in Computer Science

University of Maryland, College Park 2015-now

M.Sc. in Computer Science

University of Maryland, College Park 2015-2018

B.Eng. in Computer Science

Universidad Simón Bolívar 2009-2014 · Cum Laude

Work Experience

Facebook Inc.

Software Engineer PhD Intern Summer 2019

University of Maryland

TA & Summer Lecturer 2015-now

Venezuela Inteligente

Co-founder & Developer 2012-2017

Univ. Simón Bolívar

Research & Teaching Assist. 2012-2015

News

Sep 6, 2021 Paper accepted to ESA’21 (European Symposium on Algorithms), Lisbon, Portugal (virtual).
Jun 7, 2021 Accepted talk for the 5th Workshop on Geometry and Machine Learning. This will take place during SoCG’21 (Symposium on Comp. Geometry), Buffalo, USA (virtual).
Sep 7, 2020 Paper accepted to ESA’20 (European Symposium on Algorithms), Pisa, Italy (virtual).
Aug 5, 2020 Paper accepted to CCCG’20 (Canadian Conference in Comp. Geometry) at the University of Saskatchewan, Canada (virtual), and was awarded the Best Student Paper! :trophy:
Aug 8, 2019 Paper accepted to CCCG’19 (Canadian Conference in Comp. Geometry) at the University of Alberta, Canada. The paper was later invited to a special issue of the CGTA Journal.
Jun 20, 2019 Presented at the 4th Workshop on Geometry and Machine Learning. This took place during SoCG’19 (Symposium on Comp. Geometry), Portland, USA.

Selected Publications

  1. ESA’21
    Boundary-Sensitive Approach for Approximate Nearest-Neighbor Classification
    Flores-Velazco, Alejandro, and Mount, David M.
    29th Annual European Symposium on Algorithms, 2021
  2. ESA’20
    Coresets for the Nearest-Neighbor Rule
    Flores-Velazco, Alejandro, and Mount, David M.
    28th Annual European Symposium on Algorithms, 2020
  3. CCCG’20
    Social Distancing is Good for Points too!
    Flores-Velazco, Alejandro
    32nd Canadian Conference on Computational Geometry, 2020
  4. CCCG’19
    Guarantees on Nearest-Neighbor Condensation Heuristics
    Flores-Velazco, Alejandro, and Mount, David M.
    31nd Canadian Conference on Computational Geometry, 2019