Research · SciML
Symbolic Regression in Scientific Machine Learning: From Data Noise to Governing Equations
How symbolic regression recovers compact, interpretable mathematical structures from complex datasets — bridging data complexity and governing equations through principled algorithmic search.
April 2026
Symbolic Regression · PySR · SINDy
→
Research · Methodology
Scientific Machine Learning vs. Deep Learning: A Structural Comparison
A rigorous comparison of inductive biases, generalisation properties, and scientific applicability — and why the distinction between statistical interpolation and structural generalisation matters.
April 2026
SciML · PINN · Deep Learning
→
Drug Discovery · Methodology
QSAR Model Validation: Structural Leakage and the Limits of Standard Benchmarks
Structural leakage inflates QSAR performance metrics systematically. This article analyses the mechanism, quantifies its impact in a kinase inhibitor dataset, and presents the Honest OOF Protocol as a rigorous alternative.
April 2026
QSAR · Validation · Drug Discovery
→
Methodology · Validation
Out-of-Fold Validation: Rigorous Generalisation Assessment in Scientific Models
A technical treatment of OOF validation: its construction, advantages over naive train/test splits, implementation considerations for temporal and structural data, and its use in diagnosing model pathologies.
April 2026
OOF · Cross-Validation · Generalisation
→
Finance · Interpretability
Interpretable AI in Quantitative Finance: Regime Detection and Symbolic Models
Non-stationarity, regime transitions, and regulatory explainability requirements demand more than black-box prediction. Symbolic regression applied within a regime-aware framework offers structural insight unavailable in standard ML approaches.
April 2026
Quant Finance · Regime Detection · Interpretability
→