LoG Conference Brings New York to Jersey City

The New York meetup of the Learning on Graphs Conference (LoG NY), an annual conference that covers research broadly related to machine learning on graphs, was hosted on February 29 and March 2 by the Ying Wu College of Computing (YWCC) at their NJIT@JerseyCity location, on the premises of the NJIT-BGU Institute of Future Technologies (IFT).

The event was spearheaded by Ali Parviz, a Ph.D. student at NJIT, along with Ph.D. students from Cornell and MIT. The LoG meetups are local events held in various national and international locations to foster discussions between students, professors, and industry researchers. Their Graph Machine Learning and Geometric Deep Learning communities have been rapidly advancing the state-of-the-art in data-centric network algorithms to innovate everything, from recommending people for jobs through LinkedIn, and music on Spotify, to drug discovery.

The recent event was the first in the New York metro area and for NJIT, following previous such events in various national and international locations. There were 14 featured speakers, comprising researchers from Google, Meta, IBM, VantAI, Flatiron Institute and Genentech; and notable faculty from the Courant Institute of Mathematical Sciences, Rutgers University, Cornell University and NJIT, including Assistant Professor Mengjia Xu, a recent addition to the department of Data Science, who delivered a lecture on Hyperbolic Brain Representations for Subjective Cognitive Decline Prediction and Detecting Healthy Brain Aging Trajectories.

Other presentations focused upon a wide diversity of special topics of interest, including drawing connections between Geometric Deep Learning with other important current advances in Quantum Computing and Large Language Models (LLMs).

The event also featured a forward-looking panel discussion, led by NJIT faculty Yiannis Koutis, on the challenges and opportunities of the near future. Several panelists discussed how the next big step in advancing and democratizing AI may come from algorithmic advances that will lead to smaller LLMs and/or LLMs employing more energy-efficient computational operations, and specialized hardware. “Very few people would anticipate the dramatic advent of ChatGPT only three years ago, and as computer scientists, we need to have such open inspiring conversations about what the next breakthrough might be”, Koutis said.

Ali Parviz and the other organizers look forward to establishing the NYC LoG meetups as a permanent recurring event in the coming years. Several participants expressed their enthusiasm with NJIT’s “spectacular” and conveniently located facility where YWCC offers Graduate Certificate and MS programs in Computer Science, Data Science, and more recently Artificial Intelligence. Vincent Oria, chair of the Computer Science department at NJIT, and Baruch Schieber, director of the Institute for Future Technologies, said that YWCC and NJIT intend to provide their Jersey City facility for organizing similar events that bring together prominent researchers.