Relevance
8/10
Highly relevant to quantitative trading and portfolio management. Addresses the practical cardinality constraint that limits the number of assets in a portfolio, which is critical for real-world fund management. The multi-objective approach provides Pareto-optimal portfolios for different risk profiles. The scalability to large markets (418+ assets) makes it applicable to institutional portfolio construction. However, it focuses on static portfolio construction rather than dynamic trading strategies.
Implementation Complexity
8/10
High complexity due to: (1) novel two-row encoding requiring careful weight normalization, (2) multiple tournament selection strategies with phase-dependent activation, (3) three types of mutations with dynamic probability distributions, (4) explorer operator with five heuristic fund allocation methods and iterative search, (5) three distinct CC repair techniques (associated-based with clustering, score-based with K-means, threshold-based with correlation analysis), (6) modified BC repair handling repeated assets, (7) two-phase optimization with different strategies per phase. Requires careful parameter tuning (alpha, beta, lambda, gamma, Omega, etc.).
Reproducibility
3/5
Code and data available upon request. OR-library benchmark datasets (S&P 100, DAX 100, Nikkei 225) are publicly available. TSE data described with specific time period (March 2017 - May 2023). Algorithms implemented in Python with Numba JIT compilation. Population size, iterations, and constraint parameters specified. However, no public repository link provided.