Professor
IEEE Fellow
University Faculty Scholar

School of Industrial Engineering

School of Electrical and Computer Engineering (ECE) (by courtesy)

Purdue University
West Lafayette, IN, 47907-5400, USA
Office: 384, Grissom Hall
Email: gscutari[at]purdue[dot]edu
Fax: +1-765-494-6802
Phone: +1-765-494-7342

[Google Scholar Profile]

[CV]
Stacks Image 1466

Professional Experiences

Areas of Interest

  • Machine Learning and Algorithms
  • Mathematical Optimization Theory/Methods and (Engineering) Applications
  • Distributed Computing and Information Processing
  • Algorithmic Game Theory and Applications
  • Variational Inequality Methods for Engineering

Openings for Ph.D Students and Postdoc

I am looking for self-motivated students who have strong mathematical background, and are interested in theoretical aspects of large-scale optimization, distributed/parallel computing, and machine learning. More details on some recent works can be found here and here. Please contact me if you are interested in joining my group. Position in the School of Industrial Engineering and the School of Electrical and Computing Engineering are available.

I also have two opening for post-docs.

Postdoctoral positions on distributed optimization and/or statistical learning and/or information processing

Recent News

  • 2023: We have received the Purdue Seed of Success ACORN Award.
  • 2023: I have been selected as one of the 2023 Purdue University Faculty Scholars.
  • 2023: I received the 2023 Purdue College of Engineering Faculty Excellence Award in Research.
  • 2022: I have been selected as one of the 2023 IEEE Signal Processing Society's Distinguished Lecturers. More details here.
  • 2022-New Results: Three papers accepted to NeurIPS 2022:
    • ``Optimal Gradient Sliding and Its Application to Distributed Optimization under Similarity," with D. Kovalel, A. Beznosikov, E. Borodich, and A. Gasnikov; [PDF]
    • ``DGD^2: A Linearly Convergent Distributed Algorithm for High-dimensional Statistical Recovery," with M. Maros;
    • ``Acceleration in Distributed Sparse Regression," with M.. Maros
  • 2022-New Result: "High-Dimensional Inference over Networks: Linear Convergence and Statistical Guarantees," [PFD]. This paper studies the statistical and computational guarantees of a gradient-tracking (distributed) algorithm solving a linear regression problem over networks.
  • 2021-New Result: "Distributed Sparse Regression via Penalization" [PDF]. This work studies statistical and computation guarantees of a DGD-like (distributed) algorithms solving linear regression problems over networks.
  • 2021-New Result: "Acceleration in Distributed Optimization under Similarity" accepted for publication, AISTATS 2022 [PDF]
  • 2021-New Result: "Distributed Saddle-point under similarity'" NeurIPS 2021 [PDF]
  • 2021-New Result: "Finite-Bit Quantization for Distributed Algorithm with Linear Convergence" [PDF]
  • 2021-New Paper: "Newton Method over Networks is Faster up to the Statistical Precision," ICML 2021 [PDF]
  • Dec. 2020: I gave a talk in the "One World Signal Processing Seminar Series". The link to the talk is here.
  • 2020 IEEE SP Best paper Award: Our paper "Joint Optimization of Radio and Computational Resources for Multicell Mobile-Edge Computing" has been selected for the 2020 IEEE SPS Best Paper Award.
  • I have been elevated to IEEE Fellow (Class 2021)
  • New Results: "Distributed Optimization based on Gradient-tracking Revisited: Enhancing Convergence Rate via Surrogation" [PDF]
  • New Paper: Francisco Facchinei, Vyacheslav Kungurtsev, Lorenzo Lampariello, and Gesualdo Scutari, "Diminishing Stepsize methods for Composite Problems with Ghost Penalty: From the General to the Convex Regular Constrained case," July 2020 [PDF]
  • New Paper: Jinming Xu, Ye Tian, Ying Sun, and Gesualdo Scutari, "Distributed Algorithms for Composite Optimization: Unified and Tight Convergence Analysis," preprint, [PDF]
  • Jan 2020: New Paper: Jinming Xu, Ye Tian, Ying Sun, and Gesualdo Scutari, "Accelerated Primal-Dual Algorithms for Distributed Smooth Convex Optimization over Networks," AISTAT 2020 [PDF]
  • Oct. 2019: New Paper: Ye Tian, Ying Sun, and Gesualdo Scutari, "Asynchronous Decentralized Successive Convex Approximation,” [PDF]
  • May 2019: New Paper: Ying Sun, Amir Daneshmand, and Gesualdo Scutari, "Convergence Rate of Distributed Optimization Algorithms based on Gradient Tracking”. This paper proves for the first time linear convergence of a distributed algorithm—termed SONATA—applicable to the minimization of a (strongly convex) smooth plus a non smooth function over a network, modeled as (possibly time varying and directed) graph. [PDF]
  • January 2019: New Paper: Chang-Shen, Nicolo' Michelusi, and Gesualdo Scutari, "Limited Rate Distributed Weight-Balancing and Average Consensus over Digraphs" [PDF]
  • November 2018: New Paper: Ye Tian, Ying Sun, and Gesualdo Scutari, "Achieving Linear Convergence in Distributed Asynchronous Multi-agent Optimization". The paper proposes the first distributed algorithm over directed graphs (networks) that achieves linear convergence in a very general asynchronous setting (agent updates they variables in an uncoordinated fashion and possibly using delayed information) [PDF]
  • September 2018: New Paper: Amir Daneshmand, Gesualdo Scutari, and Vyacheslav Kungurtsev "Second Order Guarantees of Distributed Gradient Algorithms". The paper studies second order guarantees of the DGD algorithm and distributed gradient-tracking based algorithms for nonconvex optimization. We proved that DGD algorithm likely converges to a neighborhood of a second order stationary solution, and the family of distributed algorithms based on gradient tracking converge to exact second order stationary solutions for almost all initializations, avoiding thus saddle points [PDF].
  • August 2018: New Paper: Ivano Notarnicola, Ying Sun, Gesualdo Scutari, and Giuseppe Notarstefano, "Distributed Big-Data Optimization via Block-Iterative Gradient Tracking". The paper proposed a distributed algorithm for nonconvex multiagent optimization over (di)graphs unlocking for the first time block communications and optimizations from the agents' site. [PDF].
  • August 2018: New Paper: Ye Tian, Ying Sun, Gesualdo Scutari, and Bin Du, "ASY-SONATA: Achieving Linear Convergence Rate in Distributed Asynchronous Multi-agent Optimization". The paper proposed a fairly general distributed asynchronous algorithmic framework achieving linear convergence over (di)graphs [PDF].
  • August 2018: New Paper: Amir Daneshmand, Ying Sun, Gesualdo Scutari, Francisco Facchinei, and Brian Sadler, ''Decentralized Dictionary Learning over Time-varying Digraphs”. This paper introduces, analyzes, and tests numerically the first provably convergent distributed method for a fairly general class of Dictionary Learning problems [PDF].
  • May 2018: Our last effort — Gesualdo Scutari and Ying Sun, "Parallel and Distributed Successive Convex Approximation Methods for Big-Data Optimization", C.I.M.E Lecture Notes in Mathematics, Springer Verlag Series, 2018, 158 pages. [PDF]
  • April 2018: I have been awarded a Rising Star Professorship in Industrial Engineering. I'm now elevated as the Thomas and Jane Schmidt Rising Star Associate Professor.
  • Dec. 2017: The paper "Distributed Big-Data Optimization via Block Communications" by I. Notarnicola, Y. Sun, G. Scutari, G. Notarstefano, received the Best Student Paper Award at the 2017 IEEE International Workshop on Computational Advances in Multi-Sensor Adaptive Processing.
  • July 2017: New NSF Grant: CIF: Distributed Fog-Computing for Nonconvex Big-Data Analytics. Details here.
  • July 2017: I'm the keynote speaker at the ACM MobiHoc 2017 Workshop on "Distributed Information Processing in Wireless Networks", July 10-14, 2017, IIT Madras, Chennai, India.
  • July 2017: I will give the plenary talk, entitled “Just Relax: Parallel Distributed Nonconvex Optimization via Successive Convex Approximation,” at the IEEE International Workshop on Signal Processing Advances in Wireless Communications (SPAWC 2017), July 3-6, 2017, Hokkaido, Japan. Slides can be found here.
  • March 2017: I will give the invited talk “In-Network Non-Convex Large-Scale Optimization”, at the Conference on Nonconvex Statistical Learning, University of Southern California, May 26-27, 2017, Los Angeles, USA.
  • March 2017: I will give the talk “Parallel and Distributed Methods for Big-Data Optimization,” March 29, 2017, NSF BDSpoke project - Workshop on data quality, Purdue University, West-Lafayette, IN, USA.
  • Feb. 2017: “In-Network Nonconvex Large-scale Optimization,” Feb. 8, 2017, Information Systems Laboratory Colloquium, Electrical Engineering Department, Stanford University, CA, USA.
  • Feb. 2017: I will give the talk “In-Network Nonconvex Large-scale Optimization,” Feb. 7, 2017, Institute for Pure & Applied Mathematics, University of California (UCLA), Los Angeles, CA, USA.
  • November 2016: I will give the talk “In-Network Nonconvex Large-scale Optimization,” Dept. of Electrical Engineering, University of California (UCLA), Los Angeles, CA, USA.
  • October 2016: New Paper-"Feasible Methods for Nonconvex Nonsmooth Problems with Applications in Green Communications," The paper proposes a fairly general algorithmic framework based on successive convex approximation, suitable for nonconvex nonsmooth constrained optimization problems. The model and algorithms can be applied to a variety of problems from many field, including signal processing, machine learning, and communications [PDF].
  • October 2016: I will give the talk “In-Network Nonconvex Large-scale Optimization,” Dept. of Industrial Engineering and Management Science, Northwestern University, Evanston, IL, USA.
  • August 2016: I will give the talk “In-Network Nonconvex Large-scale Optimization,” the Fifth International Conference on Continuous Optimization (ICCOPT 2016), August 6-11, 2016, Tokyo, Japan.
  • July 2016: New Paper-"Asynchronous Parallel Algorithms for Nonconvex Big-Data Optimization: Model and Convergence". The paper proposes a novel asynchronous parallel algorithmic framework for nonconvex optimization problems. The proposed framework hinges on successive convex approximation techniques and a novel probabilistic model that captures key elements of modern computational architectures and asynchronous implementations in a more faithful way than current state of the art models. Part I-[PDF]
  • July 2016: I will give the talk “In-Network Nonconvex Large-scale Optimization,” July 7, 2016, Dept. of Engineering, University of Salento, Lecce, Italy.
  • June 2016: I will give the talk “In-Network Nonconvex Large-scale Optimization,” Dept. of Computer, Control, and Management Engineering Antonio Ruberti, Sapienza University of Rome, Rome, Italy.
  • April 2016: I will give the talk “Parallel Asynchronous Algorithms for Nonconvex Large-Scale Optimization,” INFORMS Optimization Society Conference, March 17-19, 2016, Princeton University.
  • March 2016: I will give the talk “Parallel Asynchronous Algorithms for Nonconvex Large-Scale Optimization,” INFORMS Optimization Society Conference, March 17-19, 2016, Princeton University.
  • March 2016: I will give the talk “In-network Nonconvex Big-data Optimization,” ISE Dept., University of Illinois at Urbana Champaign, IL, USA.
  • Feb. 2016: New Paper-“NEXT: In-network Nonconvex Optimization”. The paper studies nonconvex distributed optimization in multi-agent networks with time-varying (nonsymmetric) connectivity. We introduce the first convergent algorithmic framework for the distributed minimization of the sum of a smooth (possibly nonconvex and nonseparable) function–the agents’ sum-utility– plus a convex (possibly nonsmooth and nonseparable) regularizer. The proposed method hinges on successive convex approximation techniques while leveraging dynamic consensus as a mechanism to distribute the computation among the agents [PDF].
  • Jan. 2016: New Two-Part Paper--"Distributed Methods for Nonconvex Constraints Multi-Agent Problems-Part I & Part II: Theory & Applications in Communications and Machine Learning”. This two-part paper proposes a general distributed algorithmic framework for the minimization of a nonconvex smooth (sum-utility) function subject to nonconvex smooth constraints. Our framework is very general and flexible; it unifies several existing Successive Convex Approximation-based algorithms such as (proximal) gradient or Newton type methods, block coordinate (parallel) descent schemes, difference of convex functions methods, and improves on their convergence properties. Part II customizes our general methods to several multi-agent optimization problems, mainly in communications and networking; the result is a new class of (distributed) algorithms that compare favorably on existing ad-hoc (centralized) schemes (when they exist) [Part I] [Part II].
  • Nov. 2015: Congratulations to Peiran Song who won the 2015 IEEE Signal Processing Society Young Author Best Paper Award, for the joint paper: Gesualdo Scutari, Francisco Facchinei, Peiran Song, Daniel P. Palomar, and Jong-Shi Pang, “Decomposition by Partial Linearization: Parallel Optimization of Multiuser Systems,” IEEE Trans. on Signal Processing, vol. 63, no. 3, pp. 641-656, Feb. 2014.
  • October 2015: I received the 2015 AnnaMaria Molteni Award for Mathematics and Physics, ISSNAF, Washington DC.
  • October 2015: I will give the talk “In-Network Nonconvex Large-Scale Optimization,” at the Army Research Laboratory, Adelphy, MD, USA.
  • October 2015: I was appointed as Scientific Director for Big Data Analytics in the Cyber Center (Discovery Park, Purdue University).
  • September 2015: New NSF Grant: CIF: Parallel Online Algorithms for Large-Scale MRI. Details here.
  • August 2015: New NSF Grant: CIF:Communicating While Computing: Mobile Fog Computing Over Wireless Heterogeneous Networks. Details here.
  • July 2015: I will give the talk “In-Network Nonconvex Large-Scale Optimization,” the 2015 Simulation Workshop, Purdue University, West Lafayette, IN, USA.
  • July 2015: I will give the talk “In-Network Nonconvex Optimization,” 22nd International Symposium on Mathematical Programming, July 12-17, 2015, Pittsburgh, PA, USA.
  • May 2015: I will give the talk “Nonconvex Distributed Optimization over Networks,” 2015 IEEE Communication Theory Workshop (CTW 2015), May 10-13 2015, Orange Country, CA, USA.
  • March 2015: I will give the talk “Parallel and Distributed Algorithms for (nonconvex) Big-Data Optimization”, Dept. of Electrical and Computer Engineering, University of Wisconsin-Madison.
  • March 2015: I will give the talk “Parallel and Distributed Algorithms for (nonconvex) Big-Data Optimization”, Dept. of Industrial Engineering, Purdue University.
  • Feb. 2015: I will give the talk “Parallel and Distributed Algorithms for (nonconvex) Big-Data Optimization”, Dept. of Industrial & Operations Engineering, University of Michigan, Ann Arbor.
  • Feb. 2015: I will give the talk “Parallel and Distributed Algorithms for (nonconvex) Big-Data Optimization”, Dept. of Industrial and Manufacturing Engineering, Penn State University.
  • Dec. 2014: I’m a Guest Editor of the special issue “Signal Processing for Big Data”, EURASIP J. on Advances in Signal Processing [CFP].
  • Dec. 2014: New Paper—“Joint Optimization of Radio and Computational Resources for Multi-Cell Mobile-Edge Computing’”. This paper formulates the offloading problem as the joint optimization of the radio and computational resources in order to minimize the overall users’ energy consumption, while meeting latency constraints; centralized and decentralized algorithms with provable convergence are proposed [PDF].
  • Dec. 2014: I will give the talk "Parallel Algorithms for Big-Data Optimization,” Dept. of Information Engineering, Electronics, and Telecommunication, University of Rome “La Sapienza”, Rome, Italy.
  • Dec. 2014: I will give the talk “HyFLEXA: An Hybrid Random/Deterministic Parallel Algorithm for Nonconvex Big Data Optimization,” Dept. of Electrical Engineering and Computer Science, University of Michigan, Ann Arbor, MI, USA.
  • Dec. 2014: I will give the talk "Parallel Algorithms for Big-Data Optimization,” Dept. of Electrical and System Engineering, University of Pennsylvania (UPenn), Philadelphia, PA, USA.
  • October 2014: New Paper--"A Parallel Stochastic Optimization Method for Nonconvex Multi-Agent Optimization Problems". This paper proposed the first distributed best-response (i.e., non gradient-like) algorithmic framework for nonconvex multi-agent problems. The framework can be applied to a variety of problems in signal processing, communications, and networking [PDF].
  • July 2014: New Paper--"Hybrid Random/Deterministic Parallel Algorithms for Nonconvex Big Data Optimization". This paper develops for the first time general parallel (random/deterministic) algorithms for nonconvex nonsmooth large-scale optimization problems [PDF].
  • June 2014: I will teach a course entitled “Advanced Decomposition Algorithms for Multi-agent Systems ” at the CIME Summer School on “Centralized and Distributed Multi-agent Optimization: Models and Algorithms”, Cetraro, Italy, June 23-28, 2014. More info here; Registration is free!
  • April 2014: I will give the talk "Parallel and Distributed Optimization of Large-Scale Systems," Dept. of Electrical Engineering, Viterbi School of Engineering, University of Southern California (USC), Los Angeles, CA, USA.
  • April 2014: I will give the talk "Parallel and Distributed Optimization of Large-Scale Systems," Dept. of Electrical Engineering, Penn State University, State College, PA, USA.
  • March 2014: I will give the talk "Parallel and Distributed Optimization of Large-Scale Systems," Dept. of Electrical Engineering, University of California (UCLA), Los Angeles, CA, USA.
  • Feb. 2014: New Paper--"Parallel Algorithms for Big Data Optimization". This paper proposes a decomposition framework for the parallel (greedy) optimization of nonconvex nonsmooth functions, arising naturally in large-scale optimization problems [PDF].
  • Sept. 2013: I will visit the Dept. of Electrical and Computer Engineering and the Digital Technology Center at University of Minnesota.
  • August 2013: I will visit the Dept. of Electrical and Computer Science at the Nothwestern University.
  • August 2013: I received the 2013 UB Exceptional Scholars – Young Investigator Award.
  • August 2013: I'm an invited speaker at the IMSE summer school on Multi-Agent Networked Systems, University of Illinois at Urbana Champaign, August 15-19, 2013.
  • July 2013: I will give the invited talk "Distributed Optimization of Multiuser Systems," DIET Dept., University of Rome, "La Sapienza", Rome, Italy.
  • June 2013: I will give the invited lectures "Parallel Optimization of Large-Scale Multi-Agent Systems" in the 59th Workshop of the International School of Mathematics, entitled "Nonlinear Optimization: a Bridge from Theory to Applications," Erice, Sicily, Italy.
  • April 2013: I'm an Associated Editor of IEEE Transactions on Signal Processing.
  • Feb. 2013: I received the 2013 NSF Faculty Early Career Development (CAREER) award.
  • Dec. 2012: I will give the invited talk "Complex Variational Inequalities and Applications" in the 2012 Workshop on Complementarity and its Extensions, Institute of Mathematical Sciences, National University of Singapore, Singapore.
  • Dec. 2012: I will give the tutorial "Distributed Algorithms for Multiuser Systems" in the 2012 Workshop on Complementarity and its Extensions, Institute of Mathematical Sciences, National University of Singapore, Singapore.
  • Nov. 2012: I will give the short course "Variational Inequality Theory: A Mathematical Framework for Distributed Decision Makings" in the Electrical Engineering and Computer Science Dept. at Syracuse University, NY, USA.
  • Aug. 2012: I will give the invited talk "Monotone Communication Games" in the 21st International Symposium on Mathematical Programming (ISMP), Berlin.
  • July 2012: National Science Foundation (NSF) funds our research on "Distributed Decision Making in Cognitive Ad-Hoc Networks based on Bilevel Equilibrium Programming" (role: PI).
  • Jan. 2012: I'm an Associated Editor of IEEE Signal Processing Letters (Jan. 2012-Dec. 2014).
  • Nov. 2011: I have been elected member of the IEEE Signal Processing for Communications and Networking Technical Committee (SPCOM TC) for a three year term (Jan. 2012 - Dec. 2014).
  • Highly cited paper (ISI Web of Knowledge) status for my 2008 IEEE Trans. SP paper "Optimal Linear Precoding Strategies for Wideband Noncooperative Systems Based on Game Theory – Part I: Nash Equilibria," (coauthored by D. P. Palomar and S. Barbarossa) on Game Theory applied to the design of Multiuser Systems.
  • July 2011: I will give the tutorial "Variational Inequality Theory: A Mathematical Framework for Distributed Decision Makings" in Acropolis/COST IC0902 First International Summer School on Cognitive Wireless Communications, Florence, Italy.
  • May 2011: I will give the tutorial "Variational Inequality Theory: A Mathematical Framework for Multiuser Communication Systems and Signal Processing" in IEEE ICASPP 2011, Prague, Czech Republic (together with Daniel P. Palomar).
  • Aug. 2010: I will give the tutorial “Convex Optimization, Game Theory, and Variational Inequality Theory in Multiuser Communication Systems” in EUSIPCO 2010, Aalborg, Denmark (together with Daniel P. Palomar).
Stacks Image 130