Aaron Courville 2026-11-05
Aaron Courville coauthored Deep Learning by Ian Goodfellow and others with Ian Goodfellow and Yoshua Bengio. The book develops the mathematical and statistical foundations of learning with multilayer representations, including optimization, regularization and generative models.[ref]
He also coauthored the 2014 generative adversarial network paper. Its generator and discriminator are trained against one another, connecting generative modelling with a two-player optimization problem.[ref] The construction specifies a training objective; stable optimization and useful generated samples require additional analysis and empirical evaluation.
Albert W. Tucker Created 2026-10-14 Updated 2026-11-05
Albert W. Tucker worked in topology, game theory and nonlinear optimization. With Harold W. Kuhn he helped formulate the conditions now called the Karush-Kuhn-Tucker conditions, which extend the use of Lagrange multipliers to inequality constraints. He also supplied the familiar name and narrative for the prisoner's dilemma.[ref] That example makes a mathematical distinction memorable: individually advantageous choices can lead to an outcome both players prefer less than a coordinated alternative. The payoff inequalities, rather than the story alone, define the strategic conflict.
Figure 1.
Portrait of Albert W. Tucker
. Unknown photographer; First International Topology Conference, Moscow, 1935; ETH Library source; extraction by Jochen Burghardt. Public domain asserted by Commons in origin and life-plus-70 jurisdictions. Source usage terms, via Wikimedia Commons. Source.
Claude Berge Created 2026-10-14 Updated 2026-11-05
Claude Berge was a mathematician whose work helped establish modern graph theory, with connections to combinatorics, game theory and optimization. His books gave common mathematical language to networks arising in different applications. He also formulated the strong perfect graph conjecture, a major stimulus to research on perfect graphs.[ref] A perfect graph has equal chromatic number and clique number in every induced subgraph. The condition expresses when the obvious lower bound for graph coloring, supplied by mutually adjacent vertices, is always attainable; requiring it in every induced subgraph is essential.
Figure 1.
Portrait of Claude Berge
. Mihály Hujter (HUJTER Mihály), photograph in New York, 1985; Commons records VRT permission ticket 2025052210006236. Creative Commons Attribution-Share Alike 4.0 International., via Wikimedia Commons. Source.
David Gale Created 2026-10-18 Updated 2026-11-05
David Gale was a mathematician whose research connected optimization, game theory and mathematical economics. With Lloyd Shapley, he proved the existence of a stable matching under the classical two-sided preference model and gave an algorithm to construct one.[ref] Stability means that no unmatched pair would both prefer each other to their assigned partners. This is different from maximizing a single total score: the requirement rules out a mutually beneficial deviation. The Gale-Shapley algorithm became a foundation for applications such as assigning students to schools and doctors to residency programs.
Figure 1.
Portrait of David Gale
. George M. Bergman, photograph; Filetime, existing portrait crop. CC BY-SA 4.0, via Wikimedia Commons. Source.
Gabriele Farina Created 2026-10-14 Updated 2026-11-05
Gabriele Farina is a computer scientist who studies machine learning, optimization and game theory, especially decision-making with incomplete information. His research connects individual learning rules with the collective behaviour of interacting decision-makers.[ref]
One focus is no-regret learning: can repeated local adaptation lead to a useful game-theoretic equilibrium? The computational question matters because a multi-agent system must find strategies through available observations and calculations. His doctoral work studied learning and equilibrium computation in imperfect-information games.[ref]
Figure 1.
Portrait of Gabriele Farina
. Portrait identified on the MIT EECS profile.
Game Created 2026-10-20 Updated 2026-11-05
A mathematical game specifies participants, available choices, information and consequences of combined choices. A normal-form game records complete strategies and payoffs; an extensive-form game records the sequence of decisions and what each player knows at them.
For example, matching pennies gives each player two actions and opposing payoffs depending on whether their choices agree. Its equilibrium requires mixed strategies. A puzzle with one decision maker is often called a game informally, but strategic interaction and game theory enter when outcomes depend on several decision makers' choices. The rules determine the mathematical object; a physical board alone does not.
Harold W. Kuhn Created 2026-10-14 Updated 2026-11-05
Harold W. Kuhn connected game theory with continuous and discrete optimization. His work with Albert W. Tucker supplied the1951 nonlinear-programming formulation behind the Karush-Kuhn-Tucker conditions, while his1955 Hungarian algorithm addressed the assignment problem. He also developed extensive-form games and the relation between mixed strategies and behavioral strategies.[ref] These contributions distinguish three mathematical questions: local constrained optimality, a discrete cost-minimizing assignment, and which randomized decisions can represent a player's information-dependent plan.
Figure 1.
Portrait of Harold W. Kuhn
. Portrait identified on the Princeton University obituary. Image credit: Princeton University Office of Communications.
John von Neumann 2026-10-24
A mathematician whose work includes the minimax theorem and, with Oskar Morgenstern, the foundations of expected utility and game theory.
Figure 1.
Portrait of John von Neumann
. Portrait credited to wikispaces in the Wikimedia Commons file record. CC BY-SA 3.0. Source.
Lloyd Shapley Created 2026-10-18 Updated 2026-11-05
Lloyd Shapley was a mathematician whose work helped shape game theory and the mathematical study of matching markets. His research on stable matching with David Gale supplied an existence theorem and a constructive procedure, while the Shapley value gives a systematic way to allocate the value created by cooperation.[ref] The Shapley value averages each participant's marginal contribution over all orders in which a group might form. This makes a fairness rule precise through combinatorics, rather than selecting a division solely by bargaining strength or an arbitrary order of arrival.
Mathematical economics Created 2026-10-13 Updated 2026-10-20
Oskar Morgenstern Created 2026-10-25 Updated 2026-11-05
Figure 1.
Historical portrait crop of Oskar Morgenstern
. Anonymous photograph, University of the Andes, Colombia; existing Commons crop from its 1954–1955 bulletin group photograph. The source crop is only 48 by 74 pixels. Source.
Robbert Fokkink Created 2026-10-14 Updated 2026-11-05
Robbert Fokkink studies game theory, dynamical systems, ergodic theory and topology. His research statement highlights search games, symbolic sequences and the topology of inverse limits and tiling spaces.[ref] A search-game model makes uncertainty strategic: the best search plan depends on where a target can be hidden and what information the searcher receives. His historical research on Willem Abraham Wythoff also connects the provenance of early extremal graph theory with the study of mathematical games.[ref]
Stackelberg competition 2026-10-06
A sequential model in game theory in which a leader commits to an action before a follower observes it and chooses a best response. The leader anticipates the follower's response when optimizing. With private follower information, the leader maximizes its expected payoff over follower types. The resulting solution is a Stackelberg equilibrium.
Thomas S. Ferguson 2026-11-05
Thomas S. Ferguson works on statistics, probability theory and game theory. His UCLA page documents research on statistical decisions, stopping problems and games.[ref]
His 1973 paper introduced a Bayesian framework built on the Dirichlet process, a probability distribution over probability measures. If the prior parameter measure is , observing changes it to . This conjugate update extends the familiar idea of adding category counts to inference about an entire unknown distribution, providing a foundation for Bayesian nonparametric statistics.[ref]
Tom Wexler 2026-11-05
Tom Wexler works on approximation algorithms, game theory and networks. The author information accompanying Near-Optimal Network Design with Selfish Agents by Elliot Anshelevich and others identifies these research interests and his association with Denison University. It also verifies that this is the coauthor of the strategic network-design work.[ref]
A simple shared-cost example illustrates the kind of incentive question involved. Suppose each of two users can buy a private link costing four or use a common link costing six, split equally if both use it. Both choosing the common link is stable and costs six overall. Both choosing private links is also stable: switching alone to the common link would cost six rather than four. The latter arrangement costs eight despite having no profitable unilateral move.
The contrast motivates separate analysis of the best and worst Nash equilibria. Network-design work asks how cost-allocation rules and algorithms can find efficient stable outcomes, rather than assuming every stable outcome is socially cheapest.[ref]
Figure 1.
Tom Wexler, portrait identified by the linked author, institutional or Commons source
.