2/17/2025 Bruce Adams
ACM has named CS professor Arindam Banerjee a 2024 Fellow. He joins 24 others from universities, corporations, and research centers in Australia, Canada, Chile, China, Denmark, Germany, India, Israel, Italy, the Netherlands, Singapore, the United Kingdom, and the United States, selected by their peers for transformative contributions to computing science and technology.
Written by Bruce Adams
ACM has named CS professor Arindam Banerjee a 2024 Fellow. He joins 24 others from universities, corporations, and research centers in Australia, Canada, Chile, China, Denmark, Germany, India, Israel, Italy, the Netherlands, Singapore, the United Kingdom, and the United States selected by their peers for transformative contributions to computing science and technology.
Banerjee says, “Being named an ACM Fellow is truly a great honor. It is overwhelming to see the list of ACM fellows who have had tremendous impact in computer science and now being part of that group. It also fills me with a sense of gratitude, as none of the work I have done would have been possible without the mentors, colleagues, and students I have worked with over the years, who truly challenge you and bring out the best in you.”
Banerjee was recognized for his contributions to advances in statistical machine learning and its applications in science. He describes his research as being
“Primarily in statistical machine learning (ML) and its applications to real-world scientific problems. In ML, my core technical work has focused on problems with structure and randomness and how to develop statistically and computationally efficient algorithms in such settings. The specifics have considered clustering, generative modeling, approximate inference in Bayesian models, anomaly detection, sparse estimation, over-parameterized models including deep learning, and sequential decision-making. On the applied side, I have worked on problems in climate and environmental sciences, ecology, text analysis, recommendation systems, and finance.”
“ML has really grown over the past two decades,” he says. “I consider myself fortunate to be witnessing the growth from up close and contributing to it in some ways. For example, when some of us contributed to ‘generative AI’ about two decades back, trying to put the pieces together, we did not anticipate the explosive growth and success the area has seen in recent years. With the explosive growth of ML in terms of its use and impact over the past decade, my research interests on problems with structure and randomness have evolved and, in some ways, deepened. For example, while modern deep learning models use trillions of parameters, there is growing evidence (work by us and others) that the intrinsic dimensionality of such models are considerably lower. There will arguably be active algorithmic advances going forward that take advantage of such structure, and we plan to contribute to such advances.”
Some contemporary problems Banerjee and team have worked on include
- A project focused on uncertainty-quantified spatial extrapolation of plant functional traits like leaf nitrogen or leaf phosphorus globally
- Aspects of anomaly detection for time-series using semi-Markov models applied to aviation safety, in particular modeling issues encountered during the landing phase (descend from 10,000 ft) of commercial airlines. This work, supported by NASA, resulted in technology transfer to NASA.
- Development of TorchGeo, a PyTorch domain library for geospatial data, to enable Geospatial AI. TorchGeo is currently being used at CMU, UC Berkeley, Stanford University, MIT, TUM, EPFL, ETH Zurich, University of Oxford, Microsoft, Intel, IBM, Amazon, Meta, and other places.
Banerjee teaches the graduate course in generative AI, “essentially teaching students what's under the hood in such models.” He is President of the Society of Artificial Intelligence and Statistics, which runs the annual AISTATS international conference.
The ACM Fellows program recognizes the top 1% of the nearly 110,000 ACM professional members for their outstanding accomplishments in computing and/or outstanding service. Typically, a yearly cohort consists of around 50 individuals. Fellows must have had at least five years of professional membership within the last ten years. Banerjee and the other 2024 ACM Fellows will be recognized at the ACM Annual Awards Banquet on June 14, 2025, in San Francisco, California.
Arindam Banerjee is an Illinois Grainger Engineering professor of computer science. Arindam Banerjee holds the Founder Professorship in Engineering.