Zadovoljstvo nam je pozvati vas na 55. Fakultetski seminar koji će se održati u srijedu, 18.05.2022., s početkom u 12h, u amfiteatru A1-1.
Seminar će održati dr. sc. Ilgim Yaman, Giresun Univesity (Turska) na temu:
Portfolio optimization based on nonlinear neural network
Sažetak predavanja:
Portfolio selection problem is a very popular optimization problem in the optimization world. Harry Markowitz [1] had proposed standard portfolio optimization in 1952. In the Portfolio optimization problem main goal is minimizing the risk, while maximizing the expected return of portfolio. Because of portfolio optimization problem is an NP-hard problem, many heuristic methods were used to solve portfolio optimization method such as particle swarm optimization, ant colony optimization etc. Since portfolio optimization problem is a quadratic programming (QP) problem, in 2019, we integrate the neural network which represented in 2014 by Yan [2], in to standard portfolio optimization problem [3]. Proposed neural network is based on solving primal and dual problems simultaneously [4]. In 2021, we proposed another method to solve cardinality constrained portfolio optimization problem [4]. In this study, we combine genetic algorithm and nonlinear neural network to limited the number of assets in the portfolio. In this Erasmus mobility course, portfolio optimization and the proposed methods will be present.
References:
[1] Markowitz H., (1952), Portfolio selection, The journal of finance, 7(1):77-91
[2] Yan, Y. (2014), A new nonlinear neural network for solving QP problems, International Symposium on Neural Networks, Springer International Publishing, 347-357
[3] Yaman, I., & Dalkılıc, T. E. (2019). Portfolio selection based on a nonlinear neural network: An application on the Istanbul Stock Exchange (ISE30). Communications Faculty of Sciences University of Ankara Series A1 Mathematics and Statistics, 68(2), 1709-1723.
[4] Nyugen, K.V., (2000), A Nonlinear Network for Solving Linear Programming Problem s he title of proceeding, International Symposium on Matematical Programming, ISMP 2000, Atlanta, GA, USA
[5] Yaman, I., & Dalkılıc, T. E. (2021). A hybrid approach to cardinality constraint portfolio selection problem based on nonlinear neural network and genetic algorithm. Expert Systems with Applications, 169, 114517.
O predavačici:
Dr. sc. Ilgim Yaman je zaposlena na Odjelu za statistiku pri Faculty of Art and Science, Giresun University u Turskoj.
Životopis predavačice te popis publikacija nalaze se na web stranici: Rsc. Asst. Dr. Ilgım YAMAN | Department of Statistics (giresun.edu.tr)