You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Final Project for Optimization for Datascence, UNIPD MSc program 23/24. Uses variants of Frank-Wolfe algorithms for projection-free white-box adversarial attacks on convolutional neural networks.
This julia package addresses the membership problem for local polytopes: it constructs Bell inequalities and local models in multipartite Bell scenarios with binary outcomes.
Implementation of unconstrained and constrained convex optimization algorithms in Python, focusing on solving data science problems such as semi-supervised learning and Support Vector Machines.
This project was conducted as the final assignment for the Mathematical Optimization for Data Science course. The objective was to analyze and compare two variants of the Frank-Wolfe Method with the Projected Gradient Method in solving the Markowitz portfolio optimization problem.
This project was carried out as the final assignment for the Mathematical Optimization for Data Science course. The goal of the analysis was to compare two variants of the Frank-Wolfe Method with the Projected Gradient Method on the Markowitz portfolio optimization problem.
Code for the paper: [Wirth, E., Kera, H., and Pokutta, S. (2022). Approximate vanishing ideal computations at scale.](https://arxiv.org/abs/2207.01236)
Code for the paper: Wirth, E.S. and Pokutta, S., 2022, May. Conditional gradients for the approximately vanishing ideal. In International Conference on Artificial Intelligence and Statistics (pp. 2191-2209). PMLR.