Welcome!
I do research in computational mathematics with a focus on numerical optimization and machine learning.
I am an assistant professor in the School of Operations Research and Information Engineering (ORIE) at Cornell University.
I develop provably efficient optimization algorithms
that exploit problem structure and combine ideas from various fields.
I specialize in second-order methods and their global complexity bounds.
My research bridges the gap between optimization theory and
computational practice to enable faster training and better understanding of modern AI models.
More broadly, I work on:
- Optimization theory and algorithmic foundations of AI
- Convex and non-convex problem classes, complexity bounds
- Applications in statistics, machine learning, and scientific computing
- Scalable, distributed, and decentralized optimization
I received my PhD in 2021 at UCLouvain, Belgium, supervised by Yurii Nesterov. My thesis is New second-order and tensor methods in Convex Optimization. After that, I was a postdoctoral researcher at EPFL, Switzerland, working in the Machine Learning and Optimization Laboratory with Martin Jaggi.
Teaching
- ORIE 6365: Continuous Optimization: Algorithms and Complexity, Spring 26
Papers
Recent preprints / various:
-
Universal Reduced-Operator Method and High-Order Global Curvature Bounds.
Nikita Doikov and Yurii Nesterov, 2025 (arXiv) -
On the Complexity of Lower-Order Implementations of Higher-Order Methods.
Nikita Doikov and Geovani Nunes Grapiglia, 2025 (arXiv) -
Complexity of Minimizing Regularized Convex Quadratic Functions.
Daniel Berg Thomsen and Nikita Doikov, 2024 (arXiv)
Refereed publications:
2026
-
Gradient-Normalized Smoothness for Optimization with Approximate Hessians.
Andrei Semenov, Martin Jaggi, and Nikita Doikov, 2025 (International Conference on Learning Representations [ICLR], arXiv)
2025
-
Minimizing Quasi-Self-Concordant Functions by Gradient Regularization of Newton Method.
Nikita Doikov, 2023 (Mathematical Programming Journal [Math.Prog]: open access, arXiv) -
Lower Complexity Bounds for Minimizing Regularized Functions.
Nikita Doikov, 2022 (Optimization Letters Journal [Optim.Lett]: open access, arXiv) -
On-Device Collaborative Language Modeling via a Mixture of Generalists and Specialists.
Dongyang Fan, Bettina Messmer, Nikita Doikov, and Martin Jaggi, 2025 (International Conference on Machine Learning [ICML], arXiv) -
Improving Stochastic Cubic Newton with Momentum.
El Mahdi Chayti, Nikita Doikov, and Martin Jaggi, 2024 (International Conference on Artificial Intelligence and Statistics [AISTATS], arXiv) -
Cubic regularized subspace Newton for non-convex optimization.
Jim Zhao, Aurelien Lucchi, and Nikita Doikov, 2024 (International Conference on Artificial Intelligence and Statistics [AISTATS] (oral presentation), arXiv) -
First and zeroth-order implementations of the regularized Newton method with lazy approximated Hessians.
Nikita Doikov and Geovani Nunes Grapiglia, 2023 (Journal of Scientific Computing [JSC]: open access, arXiv)
2024
-
Spectral Preconditioning for Gradient Methods on Graded Non-convex Functions.
Nikita Doikov, Sebastian U. Stich, and Martin Jaggi, 2024 (International Conference on Machine Learning [ICML]: proceedings, arXiv) -
On Convergence of Incremental Gradient for Non-Convex Smooth Functions.
Anastasia Koloskova, Nikita Doikov, Sebastian U. Stich, and Martin Jaggi, 2023 (International Conference on Machine Learning [ICML]: proceedings, arXiv) -
Unified Convergence Theory of Stochastic and Variance-Reduced Cubic Newton Methods.
El Mahdi Chayti, Martin Jaggi, and Nikita Doikov, 2023 (Transactions on Machine Learning Research [TMLR], arXiv) -
Super-Universal Regularized Newton Method.
Nikita Doikov, Konstantin Mishchenko, and Yurii Nesterov, 2022 (SIAM Journal on Optimization [SIOPT]: open access, arXiv, code)
2023
-
Linearization Algorithms for Fully Composite Optimization.
Maria-Luiza Vladarean, Nikita Doikov, Martin Jaggi, and Nicolas Flammarion, 2023 (Conference on Learning Theory [COLT]: proceedings, arXiv) -
Polynomial Preconditioning for Gradient Methods.
Nikita Doikov and Anton Rodomanov, 2023 (International Conference on Machine Learning [ICML]: proceedings, arXiv) -
Second-order optimization with lazy Hessians.
Nikita Doikov, El Mahdi Chayti, and Martin Jaggi, 2022 (International Conference on Machine Learning [ICML] (oral presentation): proceedings, arXiv) -
Gradient Regularization of Newton Method with Bregman Distances.
Nikita Doikov and Yurii Nesterov, 2021 (Mathematical Programming Journal [Math.Prog]: open access, arXiv)
2022
-
High-Order Optimization Methods for Fully Composite Problems.
Nikita Doikov and Yurii Nesterov, 2021 (SIAM Journal on Optimization [SIOPT]: open access, arXiv) -
Affine-invariant contracting-point methods for Convex Optimization.
Nikita Doikov and Yurii Nesterov, 2020 (Mathematical Programming Journal [Math.Prog]: open access, arXiv, code)
2021
-
Local convergence of tensor methods.
Nikita Doikov and Yurii Nesterov, 2019 (Mathematical Programming Journal [Math.Prog]: open access, arXiv) -
Minimizing Uniformly Convex Functions by Cubic Regularization of Newton Method.
Nikita Doikov and Yurii Nesterov, 2019 (Journal of Optimization Theory and Applications [JOTA]: open access, arXiv)
2020
-
Convex optimization based on global lower second-order models.
Nikita Doikov and Yurii Nesterov, 2020 (Conference on Neural Information Processing Systems [NeurIPS] (oral presentation): proceedings, arXiv, code) -
Stochastic Subspace Cubic Newton Method.
Filip Hanzely, Nikita Doikov, Peter Richtárik, and Yurii Nesterov, 2020 (International Conference on Machine Learning [ICML]: proceedings, arXiv) -
Inexact Tensor Methods with Dynamic Accuracies.
Nikita Doikov and Yurii Nesterov, 2020 (International Conference on Machine Learning [ICML]: proceedings, arXiv, code) -
Contracting Proximal Methods for Smooth Convex Optimization.
Nikita Doikov and Yurii Nesterov, 2019 (SIAM Journal on Optimization [SIOPT]: open access, arXiv)
2018
-
Randomized Block Cubic Newton Method.
Nikita Doikov and Peter Richtárik, 2018 (International Conference on Machine Learning [ICML] (oral presentation): proceedings, arXiv)
Recent talks
- May 4, 2025: Cubic regularized subspace Newton for non-convex optimization, AISTATS, Phuket (slides)
[photo↓]
- April 9, 2025: Stochastic second-order optimization: global bounds, subspaces, and momentum, UCLouvain, Louvain-la-Neuve (slides)
- December 3, 2024: Stochastic second-order optimization: global bounds, subspaces, and momentum, WIAS, Berlin (slides)
- August 27, 2024: Spectral Preconditioning for Gradient Methods on Graded Non-convex Functions,
ALGOPT, Louvain-la-Neuve (slides) [photo↓]
- July 1, 2024: Minimizing quasi-self-concordant functions by gradient regularization of Newton method, EURO, Copenhagen (slides)
- June 26, 2024: Spectral Preconditioning for Gradient Methods on Graded Non-convex Functions, EUROPT, Lund (slides)
- June 20, 2024: Polynomial Preconditioning for Gradient Methods, FGS-24, Gijón (slides)
- April 9, 2024: Minimizing quasi-self-concordant functions by gradient regularization of Newton method, NOPTA, University of Antwerp (slides)
- August 25, 2023: Super-Universal Regularized Newton Method, EUROPT, Budapest (slides)
- July 20, 2023: Second-Order Optimization with Lazy Hessians, ICML, Hawaii (slides, poster, video)
[photo↓]
- July 19, 2023: Polynomial Preconditioning for Gradient Methods, ICML, Hawaii (poster)
- June 3, 2023: Second-Order Optimization with Lazy Hessians,
SIAM Conference on Optimization, Seattle (slides)
[photo↓]
- September 27, 2022: Super-Universal Regularized Newton Method, TML Laboratory, EPFL (slides)
- July 29, 2022: Affine-invariant contracting-point methods for Convex Optimization, EUROPT, Lisbon (slides)
- June 3, 2022: Second-order methods with global convergence in Convex Optimization, the research team of Panos Patrinos, KULeuven (slides)
- May 5, 2022: Optimization Methods for Fully Composite Problems, FGP-22, Porto (slides)
- February 21, 2022: Second-order methods with global convergence in Convex Optimization, MLO Laboratory, EPFL (slides)
- July 7, 2021: Local convergence of tensor methods, EUROPT, online (slides)
- March 4, 2021: Affine-invariant contracting-point methods for Convex Optimization, Symposium on Numerical Analysis and Optimization (invited by Geovani Nunes Grapiglia), UFPR, online (slides)
- October 28, 2020: Convex optimization based on global lower second-order models, NeurIPS, online (slides, poster)
- June 17, 2020: Inexact Tensor Methods with Dynamic Accuracies, ICML, online (slides, poster, video)
- October 8, 2019: Proximal Method with Contractions for Smooth Convex Optimization, ICTEAM seminar, Louvain-la-Neuve
- September 23, 2019: Proximal Method with Contractions for Smooth Convex Optimization,
Optimization and Learning for Data Science seminar
(invited by
Dmitry Grishchenko) Université Grenoble Alpes, Grenoble
(slides)
[photo↓]
- September 18, 2019: Complexity of Cubically Regularized Newton Method for Minimizing Uniformly Convex Functions, FGS-19, Nice (slides)
- August 5, 2019: Complexity of Cubically Regularized Newton Method for Minimizing Uniformly Convex Functions, ICCOPT, Berlin
- July 5, 2019: Randomized Block Cubic Newton Method,
Summer School on Optimization, Big Data and Applications, Veroli
[photo↓]
- June 28, 2019: Complexity of Cubically Regularized Newton Method for Minimizing Uniformly Convex Functions EUROPT, Glasgow
[photo↓]
- June 20, 2018: Randomized Block Cubic Newton Method, ICML, Stockholm
(slides,
poster,
video)
[photo↓]
- June 13, 2018: Randomized Block Cubic Newton Method,
X Traditional summer school on Optimization, Voronovo
[photo↓]
