Research & Publications

A selection of David Bookstaber's research across machine learning, quantitative finance, optimization, and aerospace systems, from 1993 through current preprints.

Machine Learning & AI

Tier Style Is Legible but Not Self-Known: A Behavioral Boundary on Within-Family Self-Recognition in Claude 4
2026 · preprint (arXiv submission pending)

An empirical study of whether large language models can behaviorally recognize their own writing among sibling models of different capability tiers, establishing a measurable boundary on emergent-introspection claims. Preregistered on OSF. Data and analysis released openly.

Quantitative Finance

Property Catastrophes and Equity Markets
AQR Capital Management white paper · April 2014 · contributions acknowledged

A quantitative study of reinsurance, property-catastrophe risk, and equity-market dependencies, for which David Bookstaber is acknowledged among the contributors.

Algorithms & Computer Science

Dynamic Routing of Circuits and Cells
Senior thesis · 1999

A mathematical analysis of routing algorithms for circuit-switched and ATM cell networks.

Simulated Annealing for the Traveling Salesman Problem
1997

An implementation and evaluation of simulated annealing applied to the classic combinatorial-optimization problem.

A Diploid Genetic Algorithm
1993

A description of a diploid genetic algorithm with performance samples and comparative analysis.

Aerospace & Human Factors

Unmanned Combat Aerial Vehicles
1998 · archived by Air University (Aerospace Power Chronicles) and the Defense Technical Information Center (DTIC ADA520346)

An early analysis of the costs, benefits, and technology trajectory of autonomous aerial systems.

Optimized Visual Interfaces for Pilots
1998

An analysis of the human visual-perception system and how cockpit interfaces can be optimized around it.