arbiter
JournalsPricingAccount

When Relevance Is Not Enough: Enhancing Financial Models through Redundancy-Aware Feature Selection

Mohamed Chelly et al.

The Journal of Financial Data Science2026https://doi.org/10.3905/jfds.2026.003article
AJG 1
Weight
0.50

What the paper says

This article examines the trade-off between feature relevance and redundancy in feature selection for linear regression models. The analysis focuses on two widely used feature selection paradigms, F-test and Random Forest, each implemented in two variants: a relevance-only version and a redundancy-aware extension that jointly accounts for both factors. To evaluate these methods, the authors construct a synthetic data generation framework that enables precise control over average feature relevance and redundancy. By assessing out-of-sample performance across a grid of such configurations, they identify the conditions under which redundancy-aware methods offer meaningful advantages. These findings are further validated on three real-world financial datasets. Results show that relying on relevance alone is often insufficient, and explicitly addressing redundancy improves both the predictive performance and the explanatory power of linear models.

Open paper page →

Cite this paper

https://doi.org/https://doi.org/10.3905/jfds.2026.003

Or copy a formatted citation

@article{mohamed2026,
  title        = {{When Relevance Is Not Enough: Enhancing Financial Models through Redundancy-Aware Feature Selection}},
  author       = {Mohamed Chelly et al.},
  journal      = {The Journal of Financial Data Science},
  year         = {2026},
  doi          = {https://doi.org/https://doi.org/10.3905/jfds.2026.003},
}

Paste directly into BibTeX, Zotero, or your reference manager.

Flag this paper

When Relevance Is Not Enough: Enhancing Financial Models through Redundancy-Aware Feature Selection

Flags are reviewed by the Arbiter methodology team within 5 business days.

Evidence weight

0.50

Balanced mode · F 0.40 / M 0.15 / V 0.05 / R 0.40

F · citation impact0.50 × 0.4 = 0.20
M · momentum0.50 × 0.15 = 0.07
V · venue signal0.50 × 0.05 = 0.03
R · text relevance †0.50 × 0.4 = 0.20

† Text relevance is estimated at 0.50 on the detail page — for your query’s actual relevance score, open this paper from a search result.