This implementation was created as part of a the course Machine Learning for Operations Research and Financial Engineering. In the following, we consider the Markowitz model for efficient portfolios. It is a convex nonlinear optimization problem that can be solved efficiently using convex global optimization techniques.
For this purpose, I have implemented a barrier term based method, which solves the problem described above for different betas in an approximate way (an efficient portfolio is calculated for each beta).
Finally, I visualized the efficient portfolios in a mu-sigma diagram (expected return versus standard deviation).