I recently skimmed through some class videos by Tucker Balch, Professor at Georgia Tech called "Machine Learning For Trading". One interesting part was portfolio optimisation.

That is, if you want to invest some amount of money, how do you distribute it

over a selection of assets in a way minimising risk while at the same time

being able to earn money through your investment. A portfolio is basically an allocation of the money into different options or in other words, the percentage of the total amount of money allocated to an option. Anyways, in the class as well as the course book the optimisation is performed using a

python method for constrained non-linear optimisation. The algorithm used to crunch these problems is called Sequential Least Squares Programming, which is implement using a fortran wrapper.

over a selection of assets in a way minimising risk while at the same time

being able to earn money through your investment. A portfolio is basically an allocation of the money into different options or in other words, the percentage of the total amount of money allocated to an option. Anyways, in the class as well as the course book the optimisation is performed using a

python method for constrained non-linear optimisation. The algorithm used to crunch these problems is called Sequential Least Squares Programming, which is implement using a fortran wrapper.

The original algorithm can be found in the ACM Algorithms catalog (Algorithm 733)

written by D. Kraft. The code is automatically translated using the fortran to python (f2py)

library.

written by D. Kraft. The code is automatically translated using the fortran to python (f2py)

library.

## No comments:

## Post a Comment