The report "Comparison between Gradient Descent and BDGD methods" has been written by Betti Gianmarco (ID: 2097050), Marinelli Andrea (ID: 2091700) e Rinaldi Giorgia (ID: 2092226). There are two files: a Python notebook in which code and algorithms have been implemented and tested on datasets; a pdf file explaining the work done.
andrea3425 / gradient-descent-and-bcgd-methods Goto Github PK
View Code? Open in Web Editor NEWThis work was completed as homework for the Mathematical Optimization for Data Science course. The objective of the analysis was to compare Gradient Descent and Block Coordinate Gradient Descent, which I implemented from scratch and tested on publicly available datasets.