Signal & Structure

Lorenzo Ognibeni — Research Fellow, CNR ISTC

Modeling Representation, Structure and Risk in Complex Systems

I develop representation learning and probabilistic modeling frameworks to study structure, dynamics and risk in complex systems under uncertainty.

Current research focus: latent structure, contrastive learning and geometric representations of neural systems.

Text analytics, semantic similarity and retrieval in specialized corpora.

Financial time-series models, regime dynamics and quantitative risk prototypes.

Causal inference, policy evaluation and shock-transmission analysis in institutional systems.

Modular framework for neural representation learning, contrastive sampling and latent dynamics.

CT image segmentation prototype with encoder-decoder CNNs and IoU evaluation.

Applied LSTM forecasting prototype for hospital-flow dynamics and triage time series.

Seed-based retrieval, TF-IDF scoring and word embeddings for legal terminology expansion.

Applied NLP pipeline for lexical analysis, sentiment and emotion profiling of political tweets.

State-space models for statistical arbitrage.

Replication-based pricing with Black-Scholes, NSS yield curves and credit-risk extensions.

Rolling quantile-regression framework for institutional systemic-risk contributions.

Black-Litterman, equal-risk contribution and VaR-based portfolio risk controls.

Bayesian analysis of policy uncertainty, institutional shocks and macroeconomic transmission.

Boundary-based causal design for historical concessions and local outcomes.