Title | Mathematical Learning Models โ Theory and Algorithms [electronic resource] : Proceedings of a Conference / edited by Ulrich Herkenrath, Dieter Kalin, Walter Vogel |
---|---|
Imprint | New York, NY : Springer New York, 1983 |
Connect to | http://dx.doi.org/10.1007/978-1-4612-5612-0 |
Descript | XIII, 226 p. online resource |
The Minimax Risk for the Two-Armed Bandit Problem -- Bandit Problems with Random Discounting -- Stochastic Approximation on a Bounded Convex Set -- Learning Automaton for Finite Semi-Markov Decision Processes -- A Local Asymptotic Minimax Optimality of an Adaptive Robbins-Monro Stochastic Approximation Procedure -- Dynamic Allocation Indices for Bayesian Bandits -- The Role of Dynamic Allocation Indices in the Evaluation of Suboptimal Strategies for Families of Bandit Processes -- On the Discretization Technique for Optimal Discounted Control of the Wiener Process -- Asymptotic Properties of Learning Models -- On the Infinitesimal Characterization of Monotone Stopping Problems in Continuous Time -- Numerical Investigation of the Two-Armed Bandit -- Uniform Bounds for a Dynamic Programming Model under Adaptive Control Using Exponentially Bounded Error Probabilities -- Stochastic Regression Models and Consistency of the Least Squares Identification Scheme -- Recursive Identification Techniques -- An Optimization Problem for Matrices with Application to Decision Models -- On a Class of Learning Algorithms with Symmetric Behavior under Success and Failure -- Convergence of a General Stochastic Approximation Process under Convex Constraints and Some Applications -- On Kerstingโs Theorem on Weak Convergence of Recursions -- On Continuous Time Learning Models -- Convergence of Stochastic Approximation Algorithms with Non-Additive Dependent Disturbances and Applications -- Sequential Probability Ratio Tests for Homogeneous Markov Chains -- Allocation Rules for Sequential Clinical Trials -- Non-Deterministic Modelling and its Application in Adaptive Optimal Control