Shape Constraints in Economics and Operations Research
Citations
19 citations
14 citations
14 citations
Cites background from "Shape Constraints in Economics and ..."
...tions research, statistical learning and engineering applications. In economics applications, for example, convexity/concavity arise in modeling utility and production functions, consumer preferences [37,18], among others. In some stochastic optimization problems, value functions are taken to be convex [35]. See also the works of [16,39,3] for other important applications of convex regression. Statistica...
[...]
References
37,989 citations
"Shape Constraints in Economics and ..." refers background or methods in this paper
...The SPAR algorithm relies on stochastic approximation theory and bears a strong resemblance to the Q-learning algorithm (Watkins and Dayan, 1992) in the reinforcement learning literature (Sutton and Barto, 1998)....
[...]
...…and reinforcement learning (RL) refer to a set of methodologies and algorithms for approximately solving complex sequential decision problems when the state space is large and/or parts of the system are unknown (Bertsekas and Tsitsiklis, 1996; Sutton and Barto, 1998; Bertsekas, 2012; Powell, 2011)....
[...]
...Approximate dynamic programming (ADP) and reinforcement learning (RL) refer to a set of methodologies and algorithms for approximately solving complex sequential decision problems when the state space is large and/or parts of the system are unknown (Bertsekas and Tsitsiklis, 1996; Sutton and Barto, 1998; Bertsekas, 2012; Powell, 2011)....
[...]
23,074 citations
"Shape Constraints in Economics and ..." refers background in this paper
...setting of Nascimento and Powell (2013). However, for a multi-dimensional state, Kunnumkal and Topaloglu (2008a) suggest arbitrarily selecting a dimension in which monotonicity is enforced....
[...]
...Recently, RL with deep neural networks has become popular (Mnih et al., 2015; Silver et al., 2016)....
[...]
[...]
16,118 citations
14,377 citations
12,671 citations
"Shape Constraints in Economics and ..." refers background or methods in this paper
...The ADMM algorithm has been proven to converge for two groups of variables; however, Mazumder et al. (2015) uses three groups of variables and the converge properties of ADMM with three groups is still an open question (Bertsekas, 1999). imsart-sts ver....
[...]
...three groups is still an open question (Bertsekas, 1999)....
[...]