Learning Optimal Discriminant Functions through a Cooperative Game of Automata

The problem of learning correct decision rules to minimize the probability of misclassification is a long-standing problem of supervised learning in pattern recognition. The problem of learning such optimal discriminant functions is considered for the class of problems where the statistical properties of the pattern classes are completely unknown. The problem is posed as a game with common payoff played by a team of mutually cooperating learning automata. This essentially results in a probabilistic search through the space of classifiers. The approach is inherently capable of learning discriminant functions that are nonlinear in their parameters also. A learning algorithm is presented for the team and convergence is established. It is proved that the team can obtain the optimal classifier to an arbitrary approximation. Simulation results with a few examples are presented where the team learns the optimal classifier.

A cooperative game of a pair of learning automata

A cooperative game played in a sequential manner by a pair of learning automata is investigated in this paper. The automata operate in an unknown random environment which gives a common pay-off to the automata. Necessary and sufficient conditions on the functions in the reinforcement scheme are given for absolute monotonicity which enables the expected pay-off to be monotonically increasing in any arbitrary environment. As each participating automaton operates with no information regarding the other partner, the results of the paper are relevant to decentralized control.

Differential Games of Mixed Type with Control and Stopping Times

We study a zero sum differential game of mixed type where each player uses both control and stopping times. Under certain conditions we show that the value function for this problem exists and is the unique viscosity solution of the corresponding variational inequalities. We also show the existence of saddle point equilibrium for a special case of differential game.

Bidding Dynamics of Rational Advertisers in Sponsored Search Auctions on the Web

In this paper, we address a key problem faced by advertisers in sponsored search auctions on the web: how much to bid, given the bids of the other advertisers, so as to maximize individual payoffs? Assuming the generalized second price auction as the auction mechanism, we formulate this problem in the framework of an infinite horizon alternative-move game of advertiser bidding behavior. For a sponsored search auction involving two advertisers, we characterize all the pure strategy and mixed strategy Nash equilibria. We also prove that the bid prices will lead to a Nash equilibrium, if the advertisers follow a myopic best response bidding strategy. Following this, we investigate the bidding behavior of the advertisers if they use Q-learning. We discover empirically an interesting trend that the Q-values converge even if both the advertisers learn simultaneously.

On the solution of bivariate population balance equations for aggregation: Pivotwise expansion of solution domain

The solution of a bivariate population balance equation (PBE) for aggregation of particles necessitates a large 2-d domain to be covered. A correspondingly large number of discretized equations for particle populations on pivots (representative sizes for bins) are solved, although at the end only a relatively small number of pivots are found to participate in the evolution process. In the present work, we initiate solution of the governing PBE on a small set of pivots that can represent the initial size distribution. New pivots are added to expand the computational domain in directions in which the evolving size distribution advances. A self-sufficient set of rules is developed to automate the addition of pivots, taken from an underlying X-grid formed by intersection of the lines of constant composition and constant particle mass. In order to test the robustness of the rule-set, simulations carried out with pivotwise expansion of X-grid are compared with those obtained using sufficiently large fixed X-grids for a number of composition independent and composition dependent aggregation kernels and initial conditions. The two techniques lead to identical predictions, with the former requiring only a fraction of the computational effort. The rule-set automatically reduces aggregation of particles of same composition to a 1-d problem. A midway change in the direction of expansion of domain, effected by the addition of particles of different mean composition, is captured correctly by the rule-set. The evolving shape of a computational domain carries with it the signature of the aggregation process, which can be insightful in complex and time dependent aggregation conditions. (c) 2012 Elsevier Ltd. All rights reserved.

Indigenous Knowledge for Biodiversity Conservation

Indigenous peoples with a historical continuity of resource-use practices often possess a broad knowledge base of the behavior of complex ecological systems in their own localities. This knowledge has accumulated through a long series of observations transmitted from generation to generation. Such ''diachronic'' observations can be of great value and complement the ''synchronic''observations on which western science is based. Where indigenous peoples have depended, for long periods of time, on local environments for the provision of a variety of resources, they have developed a stake in conserving, and in some cases, enhancing, biodiversity. They are aware that biological diversity is a crucial factor in generating the ecological services and natural resources on which they depend. Some indigenous groups manipulate the local landscape to augment its heterogeneity, and some have been found to be motivated to restore biodiversity in degraded landscapes. Their practices for the conservation of biodiversity were grounded in a series of rules of thumb which are apparently arrived at through a trial and error process over a long historical time period. This implies that their knowledge base is indefinite and their implementation involves an intimate relationship with the belief system. Such knowledge is difficult for western science to understand. It is vital, however, that the value of the knowledge-practice-belief complex of indigenous peoples relating to conservation of biodiversity is fully recognized if ecosystems and biodiversity are to be managed sustainably. Conserving this knowledge would be most appropriately accomplished through promoting the community-based resource-management systems of indigenous peoples.

mRegistry: a registry representation for fault diagnosis

Microsoft Windows uses the notion of registry to store all configuration information. The registry entries have associations and dependencies. For example, the paths to executables may be relative to some home directories. The registry being designed with faster access as one of the objectives does not explicitly capture these relations. In this paper, we explore a representation that captures the dependencies more explicitly using shared and unifying variables. This representation, called mRegistry exploits the tree-structured hierarchical nature of the registry, is concept-based and obtained in multiple stages. mRegistry captures intra-block, inter-block and ancestor-children dependencies (all leaf entries of a parent key in a registry put together as an entity constitute a block thereby making the block as the only child of the parent). In addition, it learns the generalized concepts of dependencies in the form of rules. We show that mRegistry has several applications: fault diagnosis, prediction, comparison, compression etc.

UNO Gets Easier for a Single Player

This work is a follow up to 2, FUN 2010], which initiated a detailed analysis of the popular game of UNO (R). We consider the solitaire version of the game, which was shown to be NP-complete. In 2], the authors also demonstrate a (O)(n)(c(2)) algorithm, where c is the number of colors across all the cards, which implies, in particular that the problem is polynomial time when the number of colors is a constant. In this work, we propose a kernelization algorithm, a consequence of which is that the problem is fixed-parameter tractable when the number of colors is treated as a parameter. This removes the exponential dependence on c and answers the question stated in 2] in the affirmative. We also introduce a natural and possibly more challenging version of UNO that we call ``All Or None UNO''. For this variant, we prove that even the single-player version is NP-complete, and we show a single-exponential FPT algorithm, along with a cubic kernel.

Determination of the magnetic moment by QCD sum rules

The magnetic moment of the Λ hyperon is calculated using the QCD sum-rule approach of Ioffe and Smilga. It is shown that μΛ has the structure μΛ=(2/3(eu+ed+4es)(eħ/2MΛc)(1+δΛ), where δΛ is small. In deriving the sum rules special attention is paid to the strange-quark mass-dependent terms and to several additional terms not considered in earlier works. These terms are now appropriately incorporated. The sum rule is analyzed using the ratio method. Using the external-field-induced susceptibilities determined earlier, we find that the calculated value of μΛ is in agreement with experiment.

Gluon-field contribution in QCD sum rules for the magnetic moments of the nucleons

The Wilson coefficient corresponding to the gluon-field strength GμνGμν is evaluated for the nucleon current correlation function in the presence of a static external electromagnetic field, using a regulator mass Λ to separate the high-momentum part of the Feynman diagrams. The magnetic-moment sum rules are analyzed by two different methods and the sensitivity of the results to variations in Λ are discussed.