Scientific Computing Laboratory
The Scientific Computing Laboratory carries out research activities in the field of numerical analysis, with a particular focus on the development of algorithms for the solution of optimization and linear algebra problems. The laboratory activities are conducted by the NODA group, whose research is devoted to the design of computational methods supported by solid theoretical foundations, which are subsequently implemented in Matlab and Python and released as public-domain software. In recent years, a significant part of the research has focused on large-scale problems arising in machine learning and data science. The laboratory is equipped with several high-performance workstations and a dual computing cluster, supporting advanced numerical experiments and high-performance computing applications.
Some software developed by the NODA group:
- Matlab codes for solving nonlinear systems with box constraints
TRESNEI: a Matlab trust-region solver for systems of nonlinear equalities and inequalities
CODOSOL: a bound-constrained nonlinear equations solver
STRSCNE: A scaled trust-region solver for constrained nonlinear equations
- Matlab code for the computation of smooth SVD and eigendecompositions (Schur, Takagi, generalized) for real or complex, symmetric and/or Hermitian matrix valued functions
Smooth eigendecomposition of real symmetric matrix function
The code is available in the Matlab Central File Exchange.
- Python code to the Multi-Task neural network training
ATE-SG: Alternate Through the Epochs Stochastic Gradient for Multi-Task Neural Networks
- Matlab code for matrix completion problems
IPLR_mc: Interior Point Low-Rank for Matrix Completion SDP reformulation
- Matlab code for multiclass data segmentation
GFW: a Greedy Frank Wolfe method for multiclass data segmentation
- Matlab code for derivative-free optimization problems
- Matlab code for Riemannian optimization problems
barzilaiborwein: a Riemannian Barzilai-Borwein method for optimization problems on manifolds
Last update
16.01.2026