David P. Kraines, Associate Professor Emeritus

Dr. Kraines contributed to the theory of homology and cohomology operations, particularly to Massey products and loop operations. Among the applications of his work has been his construction of the counterexample to the transfer conjecture of Quillen. He has also studied the variational bicomplex of Vinogradov and introduced the cohomology of quantum electrodynamics. Dr. Kraines has applied game theoretical techniques to study the evolution of cooperation. With Dr. Vivian Kraines, he introduced a stochastic learning approach, dubbed the Pavlov strategy, for the iterated Prisoner's Dilemma. They show that, in a noisy environment, agents using the Pavlov strategy may achieve a higher level of cooperation than those using Tit for Tat type strategies. Using computer simulations, dynamic systems and Markov chains, they extend their analysis to the evolution of the rate of learning in a society of Pavlov type agents. Recently, they have explored the natural selection of stochastic strategies in the simultaneous and the alternating Prisoner's Dilemma and identified several evolutionarily stable strategies.

Contact Info:

Office Location:	120 Science Drive, Durham, NC 27708, Durham, NC 27708
Email Address:
Web Page:	http://www.math.duke.edu/~dkrain

Office Hours:

send message to dkrain@math.duke.edu

Education:

Ph.D.	University of California, Berkeley	1965
M.A.	University of California, Berkeley	1963
B.A.	Oberlin College	1961

Research Interests: Algebraic Topology and Game Theory

Dr. Kraines contributed to the theory of homology and cohomology operations, particularly to Massey products and loop operations. Among the applications of his work has been his construction of the counterexample to the transfer conjecture of Quillen. He has also studied the variational bicomplex of Vinogradov and introduced the cohomology of quantum electrodynamics. Dr. Kraines has applied game theoretical techniques to study the evolution of cooperation. With Dr. Vivian Kraines, he introduced a stochastic learning approach, dubbed the Pavlov strategy, for the iterated Prisoner's Dilemma. They show that, in a noisy environment, agents using the Pavlov strategy may achieve a higher level of cooperation than those using Tit for Tat type strategies. Using computer simulations, dynamic systems and Markov chains, they extend their analysis to the evolution of the rate of learning in a society of Pavlov type agents. Recently, they have explored the natural selection of stochastic strategies in the simultaneous and the alternating Prisoner's Dilemma and identified several evolutionarily stable strategies.

Keywords:

Animals • Cooperative Behavior • Game Theory • Humans • Mathematics • Memory • Models, Genetic • Population Dynamics • Selection, Genetic

Curriculum Vitae

Undergraduate Research Supervised

• Theodore Freylinghuysen (2010 - 2012)  
• Hans Kist (2009 - 2011)  

Recent Publications   (More Publications)

1. Kraines, D; Kraines, V, The threshold of cooperation among adaptive agents: Pavlov and the stag hunt, Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), vol. 1193 (January, 2015), pp. 219-231, ISSN 0302-9743  [abs]
2. Kraines, DP; Kraines, VY, Natural selection of memory-one strategies for the iterated prisoner's dilemma., Journal of theoretical biology, vol. 203 no. 4 (April, 2000), pp. 335-355, ISSN 0022-5193 [10736212], [doi]  [abs]
3. Kraines, D; Kraines, V, Evolution of Learning among Pavlov Strategies in a Competitive Environment with Noise, Journal of Conflict Resolution, vol. 39 no. 3 (January, 1995), pp. 439-466, SAGE Publications [doi]  [abs]
4. Kraines, D; Kraines, V, Learning to cooperate with Pavlov an adaptive strategy for the iterated Prisoner's Dilemma with noise, Theory and Decision, vol. 35 no. 2 (September, 1993), pp. 107-150, Springer Nature, ISSN 0040-5833 [doi]  [abs]
5. Kraines, D; Kraines, V, Pavlov and the prisoner's dilemma, Theory and Decision, vol. 26 no. 1 (January, 1989), pp. 47-79, Springer Nature, ISSN 0040-5833 [doi]  [abs]