Work

Adaptive Metaheuristics for Higher-Moment Portfolio Optimization

Year: 2026

Type: software

Abstract: Replication package for two trajectory-adaptive metaheuristics for cardinality-constrained higher-moment portfolio optimization: (i) performance-adaptive differential evolution (PADE) with logistic mutation scaling driven by improvement velocity, and (ii) utility-driven particle swarm optimisation with inertia feedback based on swarm utility dispersion. Both mechanisms eliminate manual parameter tuning and are validated on a dynamic, non-convex higher-moment portfolio problem with rigorous non-parametric statistical testing (Wilcoxon, Friedman–Nemenyi) across 15 independent random seeds on S&P 500 (2005–2023), Euro Stoxx 50 (2010–2024), and BIST-100 (2010–2023) constituents.

Source: Zenodo (CERN European Organization for Nuclear Research)

DOI: 10.5281/zenodo.21070308

Author: Ouaar, Fatima

Institution: University of Biskra

Language: English

Topic: Risk and Portfolio Optimization

Subfield: Management Science and Operations Research

Field: Decision Sciences

Domain: Social Sciences

Open Access status: green

OpenAlex ID: W7166647402

Cites:

Cited by:

Related to: 10