## Farhad Pourkamali Anaraki, PhD

### Assistant Professor of Computer Science

### Director of Computational and Statistical Data Science (CSDS) Lab

### University of Massachusetts Lowell

Farhad Pourkamali Anaraki is an Assistant Professor in the Department of Computer Science at the University of Massachusetts (UMass) Lowell since fall 2018. He is the director of the Computational and Statistical Data Science (CSDS) Lab. Before joining UMass Lowell, Prof. Pourkamali spent one year as a postdoctoral research associate in the Department of Applied Mathematics at the University of Colorado Boulder, and he received his PhD in Electrical Engineering from the same institution.

## Research Areas

Development and use of artificial intelligence and machine learning in the context of computational decision support for complex systems (e.g., advanced manufacturing, structural engineering, autonomous systems)

Scalable machine learning and optimization methods (e.g., scalable clustering and data summarization)

Statistical aspects of machine learning algorithms (e.g., proving guarantees on performance)

Generative modeling and neural networks (e.g., variational autoencoders, feature extraction, classification)

Graph partitioning and approximation algorithms (e.g., spectral clustering)

Randomized numerical linear algebra (e.g., randomized SVD and PCA)

## Sponsors

Research in CSDS Lab is supported by the following organizations:

## Professional Activities and Service

**Associate Editor**of Big Data (Please submit your research!)Panelist (2x), National Science Foundation (NSF) Proposal Review

Reviewer: ICML, AISTATS, IEEE Transactions on Signal Processing, etc.

Faculty Senate Member, UMass Lowell

Faculty Senate Graduate Policy and Affairs Committee (GPAC) Member, UMass Lowell

## What's New

New paper accepted in IEEE International Conference on Machine Learning and Applications (ICMLA): "Kernel Ridge Regression Using Importance Sampling with Application to Seismic Response Prediction." Congratulations to Lydia! The acceptance rate for IEEE ICMLA 2020 is 25%, and our paper is selected for the full ORAL presentation.

Parisa Hajibabaee joined our lab in fall 2020 as a graduate research assistant.

New journal paper accepted in IEEE Robotics and Automation Letters: "A Unified NMPC Scheme for MAVs Navigation with 3D Collision Avoidance under Position Uncertainty" in collaboration with Luleå University of Technology and Caltech.

New preprint on "scalable spectral clustering using the Nystrom approximation" with new theoretical guarantees regarding the perturbation analysis of the graph Laplacian: https://arxiv.org/abs/2006.14470

New National Science Foundation (NSF) Grant on developing data-driven approaches to tailor properties of additively manufactured semi-crystalline polymers.

Guest editor of "Soft Computing and Machine Learning in Dam Engineering": https://www.mdpi.com/journal/water/special_issues/SCML

New paper accepted in MED 2020: "Unsupervised Learning for Subterranean Junction Recognition Based on 2D Point Cloud" in collaboration with Luleå University of Technology and Caltech.

Invited talk: Numerical Methods and Applied Math Seminar, Mathematical Sciences, WPI, April 9, 2020.

Invited talk: Computational and Applied Math Seminar, Department of Mathematics, Tufts University, March 9, 2020.

Prof. Pourkamali is organizing a seminar on Machine Learning in fall 2019, every Wednesday 4:30 to 5:30 PM in DAN 321. This seminar is an excellent opportunity for interested students and faculty members at UMass Lowell. For more information, you can see this page.

Prof. Pourkamali will serve as a session chair at the IEEE Machine Learning for Signal Processing Workshop (MLSP). The session is Poster Session 2, Signal Detection, Pattern Recognition, Semi/Un-supervised Learning, at 16:00-18:30pm on Monday, October 14, 2019.

Prof. Pourkamali is a program committee member for the 23rd International Conference on Artificial Intelligence and Statistics (AISTATS).

New paper accepted to Neurocomputing: "Improved fixed-rank Nystrom approximation via QR decomposition: Practical and theoretical aspects." The paper is available online here, July 2019.

New paper accepted to IEEE Machine Learning for Signal Processing (MLSP): "Large-scale sparse subspace clustering using landmarks." The paper is available on arxiv here, August 2019.

Received travel award to attend the National Science Foundation (NSF) Grants Conference, 2019.