OPTIMIZATION FOR DATA SCIENCE

Second cycle degree in DATA SCIENCE

Campus: PADOVA

Language: English

Teaching period: Second Semester

Lecturer: FRANCESCO RINALDI

Number of ECTS credits allocated: 6


Syllabus
Prerequisites: Basic knowledge of
- Real Analysis and Calculus;
- Linear Algebra.
Examination methods: - Written exam
- Project (Optional)
Course unit contents: 1. Linear optimization: Theory and algorithms
(a) Lp models for Data science;
(b) Duality (Farkas);
(c) Simplex method;
(d) Interior point methods;

2. Convex sets and convex functions
(a) Convexity: basic notions;
(c) Convex functions: Basic notions and properties (gradients, Hessians..);

3. Unconstrained convex optimization
(a) Models in data science;
(b) Characterizations of optimal sets;
(c) Gradient-type methods;

4. Constrained convex optimization
(a) Models in data science;
(b) Characterizations of optimal sets;
(c) Poyhedral approximation methods;
(d) Gradient projection methods;

5. Large scale network optimization
(a) Network models in data science;
(b) Clustering methods.