Abstract
This work chronicles research into the solution of portfolio problems with metaheuristic solvers. In particular, a genetic algorithm for solving the cardinality constrained portfolio optimisation problem with minimum asset proportions is presented and tested on the datasets of [1]. These datasets form benchmark instances used to test portfolio optimisers and are based upon indices ranging from 31 to 225 assets. The results of the GA are indicatively compared to solutions of [2] for a variety of minimum proportions, suggesting that solutions exhibit certain clustering characteristics for higher proportions. Further work is also discussed. This research is based upon the first part of the ongoing PhD thesis of the first author.
Original language | English |
---|---|
Pages (from-to) | 358-359 |
Number of pages | 0 |
Journal | Default journal |
Volume | 0 |
Issue number | 0 |
DOIs | |
Publication status | Published - 16 Jul 2019 |
Event | GECCO 2019 - Duration: 13 Jul 2019 → 17 Jul 2019 |