« DIMACS Workshop on Efficient Algorithms for High Dimensional Metrics: New Tools
May 06, 2024 - May 09, 2024
Location:
DIMACS Center
Rutgers University
CoRE Building
96 Frelinghuysen Road
Piscataway, NJ 08854
Click here for map.
Organizer(s):
Alexandr Andoni, Columbia University
Michal Koucký, Charles University
Barna Saha, University of California, San Diego
Mike Saks, Rutgers University
There are many open questions concerning algorithms related to distance measures that are not given by a norm, such as edit distance, Ulam distance, earth mover distance (Wasserstein metric), and Frechet distance. For each of these measures there are substantial gaps in our understanding of fundamental algorithmic problems such as computing, sketching, and nearest neighbor search.
The goal of this workshop is to explore and develop new tools relevant to these algorithmic problems. A prominent example of such a generic tool is metric embeddings, which map a complicated metric space into an algorithmically simpler space while preserving all distances within some small distortion factor. Such embeddings are a key tool for algorithmic applications and data structure design, e.g., for the nearest neighbor search.
For each of the aforementioned metrics there are longstanding open questions regarding the embeddings into simple spaces that minimize the worst case distortion. Recently researchers have considered relaxed notions of embeddings such as the average distortion embeddings and sketching that raise a host of new questions. Initial results on these questions have yielded significant algorithmic successes for some applications.
This workshop will bring together researchers working in computational aspects of distance measures including sketching, data structures, nearest neighbor search, string problems and communication complexity to stimulate further progress in the field. The workshop will consist of several tutorials on recent advances in the field and contributed talks. We plan for ample time for discussions in smaller groups.
Confirmed tutorial speakers:
Monday, May 6, 2024
Breakfast and Registration
Welcome by organizers
Average Distortion Embeddings and Applications to Near Neighbor Search - Part I
Aleksandar Nikolov, University of Toronto
Break (15 minutes)
Average Distortion Embeddings and Applications to Near Neighbor Search - Part II
Aleksandar Nikolov, University of Toronto
Welcome by DIMACS
Worst-case Performance of Popular Approximate Nearest Neighbor Search Implementations: Guarantees and Limitations
Piotr Indyk, Massachusetts Institute of Technology
Data-Dependent LSH for the Earth Mover's Distance
Erik Waingarten, University of Pennsylvania
Probabilistic Embedding in MPC
MohammadTaghi Hajiaghayi, University of Maryland
Break (15 minutes)
Open Problems Session
Tuesday, May 7, 2024
Breakfast and Registration
Data Structures for Fréchet Queries
Anne Driemel, University of Bonn
Break (30 minutes)
Clustering Curves under the Fréchet Distance
Anne Driemel, University of Bonn
Efficient Approximation Algorithms for Optimal Transport in Geometric Settings
Pankaj Agarwal, Duke University
Novel Properties of Hierarchical Probabilistic Partitions and their Algorithmic Applications
Alon Hovav, Hebrew University of Jerusalem
Break (30 minutes)
Diptarka Chakraborty, National University of Singapore
Wednesday, May 8, 2024
Breakfast and Registration
Dynamic Algorithms for Edit Distance - Part I
Tomasz Kociumaka, Max Planck Institute for Informatics
Break (30 minutes)
Dynamic Algorithms for Edit Distance - Part II
Tomasz Kociumaka, Max Planck Institute for Informatics
Embedding Edit Distance into Hamming Distance
Elazar Goldenberg, Academic College of Tel Aviv-Yaffo
Explicit Good Codes Approaching Distance 1 in Ulam Metric
Mursalin Habib, Rutgers University
Break (30 minutes)
Locally Consistent Decomposition of Strings with Applications to Edit Distance Sketching
Sudatta Bhattacharya, Charles University
Recovery from Non-Decomposable Distance Oracles
David P. Woodruff, Carnegie Mellon University
Thursday, May 9, 2024
Breakfast and Registration
Arnold Filtser, Bar-Ilan University
Streaming Facility Location in High Dimension
Pavel Vesely, Charles University
Fully Scalable MPC Algorithms for Clustering in High Dimension
Robert Krauthgamer, Weizmann Institute of Science
Break (15 minutes)
Light Spanners in High Dimension
Ofer Neiman, Ben Gurion University
Faster Sublinear-Time Edit Distance
Nick Fischer, Weizmann Institute of Science
Discussions
Presentations at this workshop are by invitation but others are welcome to attend. There is no fee to attend but registration is required. Please register using the button at the bottom of the page. Space is limited, so please register early if you plan to attend.
Update [March 18, 2024]: The workshop registration has been closed because we are at capacity. If you would like to join a waiting list, please send email to workshop@dimacs.rutgers.edu.
Request support: There are limited funds available to support travel by those whose attendance is contingent on support. We encourage diverse and inclusive participation and will prioritize applications for support from students and postdocs, especially those from minority or underrepresented groups. To apply for travel support please complete this form by March 15, 2024 and do not book your tickets until you hear from us! Since funds are limited, applying early provides the best access to funds.
Update [March 16, 2024]: The application for travel support is now closed.
Parking: If you do not have a Rutgers parking permit and you plan to drive to the workshop, there will be free parking in Lot 64, which is adjacent to the CoRE Building, but you must register your car to park. A link to register for parking will be provided in the confirmation message you receive when you register for the workshop.
Restaurants: For those staying in New Brunswick, there is a wide variety of restaurants within walking distance of the Heldrich and Hyatt hotels. Here are a few options.
The workshop is presented in cooperation with project CoSP, funded by European Union’s Horizon 2020 research and innovation programme under the Marie SkÅ‚odowska-Curie grant agreement No 823748.
Presented in association with the Special Focus on Lower Bounds in Computational Complexity.