Predictability

Market greenness predicts liquidity shocks tied to ESG investor preferences, and ESG-related liquidity better explains stock returns than standard measures during 2015-2019.

Theoretical analysis explains why climate foundation models succeed at field prediction but fail at extreme event detection through in-context learning complexity.

AI-driven hybrid architecture for multi-cloud storage using telemetry-guided dynamic fragmentation, encryption, and distributed trust to resist data reconstruction attacks and provider breaches.

Time-frequency analysis of Southern Oscillation Index spillovers to grain futures reveals regional asymmetries in climate vulnerability, with SAFEX maize showing significant SOI predictive capacity.

Wind shear and soil moisture interact to enhance rapid thunderstorm growth, offering new predictability for severe convective initiation across Africa and beyond.

Engineering optimization of MLOps processes for high-load recommendation systems integrating streaming features, parameter servers, and online training for latency and quality under scale.