Frank-Wolfe Variants for Portfolio Optimization - ODS23 Final Project
This file provides the necessary information to understand and use our project. This folder includes a Jupyter notebook (.ipynb) with the code, a folder containing the datasets, and a PDF report that provides a detailed explanation of the work done.
Folder Contents
- markowitz_portfolio.ipynb: This file contains the source code where we have implemented two variants of the Frank-Wolfe algorithm and the Projected Gradient Method to solve the Markowitz portfolio problem. You can also find the execution of tests on two datasets. The notebook uses the following Python libraries: NumPy, SciPy, Matplotlib
- datasets: This folder contains the two datasets used in the notebook.ipynb file. Ensure that the data is correctly placed in this folder for notebook execution.
- ODS23_project_Rinaldi_Marinelli.pdf: The report provides a detailed explanation of our work. It includes an in-depth description of the implemented algorithm, explaining how they were implemented in the .ipynb notebook. Additionally, the report covers the mathematical theory behind the algorithms and presents the results obtained from testing on the datasets.