Yadati Narahari - Academia.edu (original) (raw)

Papers by Yadati Narahari

Lecture Notes in Computer Science, 2004

Combinatorial exchanges are double sided marketplaces with multiple sellers and multiple buyers t... more Combinatorial exchanges are double sided marketplaces with multiple sellers and multiple buyers trading with the help of combinatorial bids. The allocation and other associated problems in such exchanges are known to be among the hardest to solve among all economic mechanisms. In this paper, we study combinatorial exchanges where (1) the demand can be aggregated, for example, a procurement exchange or (2) the supply can be aggregated, for example, an exchange selling excess inventory. We show that the allocation problem in such exchanges can be solved efficiently through decomposition when buyers and sellers are single minded. The proposed approach decomposes the problem into two stages: a forward or a reverse combinatorial auction (stage 1) and an assignment problem (stage 2). The assignment problem in Stage 2 can be solved optimally in polynomial time and thus these exchanges have computational complexity equivalent to that of one sided combinatorial auctions. Through extensive numerical experiments, we show that our approach produces high quality solutions and is computationally efficient.

Sadhana, 2005

Combinatorial auctions (CAs) have recently generated significant interest as an automated mechani... more Combinatorial auctions (CAs) have recently generated significant interest as an automated mechanism for buying and selling bundles of goods. They are proving to be extremely useful in numerous e-business applications such as eselling, e-procurement, e-logistics, and B2B exchanges. In this article, we introduce combinatorial auctions and bring out important issues in the design of combinatorial auctions. We also highlight important contributions in current research in this area. This survey emphasizes combinatorial auctions as applied to electronic business situations.

2008 IEEE International Conference on Automation Science and Engineering, 2008

A business cluster is a co-located group of micro, small, medium scale enterprises. Such firms ca... more A business cluster is a co-located group of micro, small, medium scale enterprises. Such firms can benefit significantly from their co-location through shared infrastructure and shared services. Cost sharing becomes an important issue in such sharing arrangements especially when the firms exhibit strategic behavior. There are many cost sharing methods and mechanisms proposed in the literature based on game theoretic foundations. These mechanisms satisfy a variety of efficiency and fairness properties such as allocative efficiency, budget balance, individual rationality, consumer sovereignty, strategyproofness, and group strategyproofness. In this paper, we motivate the problem of cost sharing in a business cluster with strategic firms and illustrate different cost sharing mechanisms through the example of a cluster of firms sharing a logistics service. Next we look into the problem of a business cluster sharing ICT (information and communication technologies) infrastructure and expl...

Sadhana, 1994

Recently, Brownian networks have emerged as an effective stochastic model to approximate multicIa... more Recently, Brownian networks have emerged as an effective stochastic model to approximate multicIass queueing networks with dynamic scheduling capability, under conditions of balanced heavy loading. This paper is a tutorial introduction to dynamic scheduling in manufacturing systems using Brownian networks.. The article starts with motivational examples. It then provides a review of relevant weak convergence concepts, followed by a description of the limiting behaviour of queueing systems under heavy traffic. The Brownian approximation procedure is discussed in detail and generic case studies are provided to illustrate the procedure and demonstrate its effectiveness. This paper places empha~sis only on the results and aspires to provide the reader with an up-to-date understanding of dynamic scheduling based on Brownian approximations.

Annals of Operations Research, 2006

In this paper, we use reinforcement learning (RL) techniques to determine dynamic prices in an el... more In this paper, we use reinforcement learning (RL) techniques to determine dynamic prices in an electronic monopolistic retail market. The market that we consider consists of two natural segments of customers, captives and shoppers. Captives are mature, loyal buyers whereas the shoppers are more price sensitive and are attracted by sales promotions and volume discounts. The seller is the learning agent in the system and uses RL to learn from the environment. Under (reasonable) assumptions about the arrival process of customers, inventory replenishment policy, and replenishment lead time distribution, the system becomes a Markov decision process thus enabling the use of a wide spectrum of learning algorithms. In this paper, we use the Q-learning algorithm for RL to arrive at optimal dynamic prices that optimize the seller's performance metric (either long term discounted profit or long run average profit per unit time). Our model and methodology can also be used to compute optimal reorder quantity and optimal reorder point for the inventory policy followed by the seller and to compute the optimal volume discounts to be offered to the shoppers.

Lecture Notes in Computer Science, 2005

This paper takes the first steps towards designing incentive compatible mechanisms for hierarchic... more This paper takes the first steps towards designing incentive compatible mechanisms for hierarchical decision making problems involving selfish agents. We call these Stackelberg problems. These are problems where the decisions or actions in successive layers of the hierarchy are taken in a sequential way while decisions or actions within each layer are taken in a simultaneous manner. There are many immediate applications of these problems in distributed computing, grid computing, network routing, ad hoc networks, electronic commerce, and distributed artificial intelligence. We consider a special class of Stackelberg problems called SLRF (Single Leader Rest Followers) problems and investigate the design of incentive compatible mechanisms for these problems. In developing our approach, we are guided by the classical theory of mechanism design. To illustrate the design of incentive compatible mechanisms for Stackelberg problems, we consider first-price and second-price electronic procurement auctions with reserve prices. Using the proposed framework, we derive some interesting results regarding incentive compatibility of these two mechanisms.

Review of Economic Design, 2015

We consider an infinite horizon dynamic mechanism design problem with interdependent valuations. ... more We consider an infinite horizon dynamic mechanism design problem with interdependent valuations. In this setting the type of each agent is assumed to be evolving according to a first order Markov process and is independent of the types of other agents. However, the valuation of an agent can depend on the types of other agents, which makes the problem fall into an interdependent valuation setting. Designing truthful mechanisms in this setting is non-trivial in view of an impossibility result which says that for interdependent valuations, any efficient and ex-post incentive compatible mechanism must be a constant mechanism, even in a static setting. Mezzetti (2004) circumvents this problem by splitting the decisions of allocation and payment into two stages. However, Mezzetti's result is limited to a static setting and moreover in the second stage of that mechanism, agents are weakly indifferent about reporting their valuations truthfully. This paper provides a first attempt at designing a dynamic mechanism which is efficient, strict ex-post incentive compatible and ex-post individually rational in a setting with interdependent values and Markovian type evolution.

arXiv preprint arXiv:1202.3751, Feb 14, 2012

Abstract: The assignment of tasks to multiple resources becomes an interesting game theoretic pro... more Abstract: The assignment of tasks to multiple resources becomes an interesting game theoretic problem, when both the task owner and the resources are strategic. In the classical, nonstrategic setting, where the states of the tasks and resources are observable by the controller, this problem is that of finding an optimal policy for a Markov decision process (MDP). When the states are held by strategic agents, the problem of an efficient task allocation extends beyond that of solving an MDP and becomes that of designing a ...

Consider a requester who wishes to crowdsource a series of identical binary labeling tasks to a p... more Consider a requester who wishes to crowdsource a series of identical binary labeling tasks to a pool of workers so as to achieve an assured accuracy for each task, in a cost optimal way. The workers are heterogeneous with unknown but fixed qualities and their costs are private. The problem is to select for each task an optimal subset of workers so that the outcome obtained after aggregating the labels from the selected workers guarantees a target accuracy level. The problem is a challenging one even in a non strategic setting since the accuracy of aggregated label depends on unknown qualities. We develop a novel multi-armed bandit (MAB) mechanism for solving this problem. First, we propose a framework, Assured Accuracy Bandit (AAB), which leads to a MAB algorithm, Constrained Confidence Bound for a Non Strategic setting (CCB-NS). We derive an upper bound on the number of time steps the algorithm chooses a suboptimal set that depends on the target accuracy level and true qualities. A more challenging situation arises when the requester not only has to learn the qualities of the workers but also elicit their true costs. We modify the CCB-NS algorithm to obtain an adaptive exploration separated algorithm which we call Constrained Confidence Bound for a Strategic setting (CCB-S). CCB-S algorithm produces an ex-post monotone allocation rule and thus can be transformed into an ex-post incentive compatible and ex-post individually rational mechanism that learns the qualities of the workers and guarantees a given target accuracy level in a cost optimal way. We also provide a lower bound on the number of times any algorithm should select a sub-optimal set and we see that the lower bound matches our upper bound upto a constant factor. We provide insights on the practical implementation of this framework through an illustrative example and we show the efficacy of our algorithms through simulations.

Sadhana, 2005

Dynamic pricing is the dynamic adjustment of prices to consumers depending upon the value these c... more Dynamic pricing is the dynamic adjustment of prices to consumers depending upon the value these customers attribute to a product or service. Today's digital economy is ready for dynamic pricing; however recent research has shown that the prices will have to be adjusted in fairly sophisticated ways, based on sound mathematical models, to derive the benefits of dynamic pricing. This article attempts to survey different models that have been used in dynamic pricing. We first motivate dynamic pricing and present underlying concepts, with several examples, and explain conditions under which dynamic pricing is likely to succeed. We then bring out the role of models in computing dynamic prices. The models surveyed include inventory-based models, data-driven models, auctions, and machine learning. We present a detailed example of an e-business market to show the use of reinforcement learning in dynamic pricing.

Annals of Mathematics and Artificial Intelligence, 2019

We study the problem of a buyer (aka auctioneer) who gains stochastic rewards by procuring multip... more We study the problem of a buyer (aka auctioneer) who gains stochastic rewards by procuring multiple units of a service or item from a pool of heterogeneous strategic agents. The reward obtained for a single unit from an allocated agent depends on the inherent quality of the agent; the agent's quality is fixed but unknown. Each agent can only supply a limited number of units (capacity of the agent). The costs incurred per unit and capacities are private information of the agents. The auctioneer is required to elicit costs as well as capacities (making the mechanism design bidimensional) and further, learn the qualities of the agents as well, with a view to maximize her utility. Motivated by this, we design a bidimensional multi-armed bandit procurement auction that seeks to maximize the expected utility of the auctioneer subject to incentive compatibility and individual rationality while simultaneously learning the unknown qualities of the agents. We first assume that the qualities are known and propose an optimal, truthful mechanism 2D-OPT for the auctioneer to elicit costs and capacities. Next, in order to learn the qualities of the agents in addition, we provide sufficient conditions for a learning algorithm to be Bayesian incentive compatible and individually rational. We finally design a novel learning mechanism, 2D-UCB that is stochastic Bayesian incentive compatible and individually rational.

Proceedings 2000 ICRA. Millennium Conference. IEEE International Conference on Robotics and Automation. Symposia Proceedings (Cat. No.00CH37065)

A supply chain network can be viewed as a network of facilities in which a customer order will fl... more A supply chain network can be viewed as a network of facilities in which a customer order will flow through internal business processes such as procurement, production, and transportation, ultimately reaching the required products on time to customers. The delivery performance of such a network can be maximized by pushing the work through the system in a way that the finished products reach the customer in a customer specified delivery window, with a very high probability. This entails synchronization and hence strict control of variability of deliveries at all intermediate points, to ensure that the m w materials and semi-finished work arrive at work spots in a timely fashion. In this paper, we explore the use of the Motorola six sigma tolemncang methodology to achieve synchronization an supply chains. In particular, we use the six sigma approach to: (1) analyze a given supply chain process f o r six sigma delivery performance; and (2) design synchronized supply chains to guarantee six sigma delivery performance. W e provide an example of a plastics industry supply chain, for which we report analysis and design experiments that demonstrate the use of the six sigma approach in designing synchronized supply chains with high levels of delivery performance.

Proceedings of the 19th ACM international conference on Information and knowledge management - CIKM '10, 2010

The advent of large scale online social networks has resulted in a spurt of studies on the user p... more The advent of large scale online social networks has resulted in a spurt of studies on the user participation in the networks. We consider a query incentive model on social networks, where user's queries are answered through her friendship network and there are `rewards' or `incentives' in the system to answer the queries utilizing ones community. We model the friendship network as a random graph with power-law degree distribution, and show that the reward function exhibits a theoretic threshold behavior on the scaling exponent α, a network parameter. Specifically, there exists a threshold on α above which the reward is exponential in the average path length in the network and below the threshold, the reward is proportional to the average path length. We demonstrate this finding on simulated power-law networks and real world data gathered from online social media such as Flickr, Digg, YouTube and Yahoo! Answers.

International Series in Operations Research & Management Science, 2011

In this chapter, we are interested in a procurement network formation problem. We present a case ... more In this chapter, we are interested in a procurement network formation problem. We present a case for modelling the procurement network formation problem as a shortest path cooperative game. We investigate recent results in shortest path cooperative games and their implications to the procurement network formation problem. We then enhance the model for procurement network formation by incorporating asymmetry in the information that agents have. Specifically, we model the procurement network formation problem as a shortest path cooperative game with incomplete information. We point out the incentive compatible core as an appropriate solution concept for this category of games. We then review the current state of the art on the topic of incentive compatible core, pose a conjecture and end with some directions for future work.

Proceedings 2002 IEEE International Conference on Robotics and Automation (Cat. No.02CH37292)

Procurement, i.e., the process of obtaining materials or services and managing their inflow into ... more Procurement, i.e., the process of obtaining materials or services and managing their inflow into organizations, is a critical process in supply chain management. Internet technologies offer attractive opportunities to reduce purchasing costs while increasing the speed and efficiency of the procurement process through seamless network integration and automated execution of key tasks. This paper attempts to model e-procurement from a lead time perspective, to capture the effects of Internet technologies on the procurement lead time. The model we develop is a multiclass queueing network called a probabilistic reentrant line. The model captures several important facets of an eprocurement process, such as use of electronic RFQ (request for quotation) in systematic sourcing, use of Internet auctions in spot purchasing, contention for procurement resources, randomness in execution times of procurement tasks, and concurrent execution of multiple procurement activities. The analysis of the queueing model suggests that end-to-end lead times in eprocurement can be compressed in ingenious ways using standard queueing theoretic techniques.

Proceedings 2003 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS 2003) (Cat. No.03CH37453)

Ahfmcl-In this paper, we come up with an innovative approach through which variability reduction ... more Ahfmcl-In this paper, we come up with an innovative approach through which variability reduction and synchro-Lead time compression in supply chains is the subject of several recent vaoers, see for examde, Gare, Narahari,. .. nimtion can he realized in supply chains. The approach and nswanaham [I], Statistical design toler&ing is a developed is founded on a connection between mechanical design tolerancing and supply chain lead time compmsion, We use two metrics for delivew oerfnrmance. deliverv sham-mature subject in the design community. The key ideas in statistical design tolerancing which provide the core

Advances in Knowledge Discovery and Data Mining, 2010

Game theory is replete with brilliant solution concepts such as the Nash equilibrium, the core, t... more Game theory is replete with brilliant solution concepts such as the Nash equilibrium, the core, the Shapley value, etc. These solution concepts and their extensions are finding widespread use in solving several fundamental problems in knowledge discovery and data mining. The problems include clustering, classification, discovering influential nodes, social network analysis, etc. The first part of the talk will present the conceptual underpinnings underlying the use of game theoretic techniques in such problem solving. The second part of the talk will delve into two problems where we have recently obtained some interesting results: (a) Discovering influential nodes in social networks using the Shapley value, and (b) Identifying topologies of strategically formed social networks using a game theoretic approach.

The 9th IEEE International Conference on E-Commerce Technology and The 4th IEEE International Conference on Enterprise Computing, E-Commerce and E-Services (CEC-EEE 2007), 2007

Formation of high value procurement networks involves a bottom-up assembly of complex production,... more Formation of high value procurement networks involves a bottom-up assembly of complex production, assembly, and exchange relationships through supplier selection and contracting decisions, where suppliers are intelligent and rational agents who act strategically. In this paper we address the problem of forming procurement networks for items with value adding stages that are linearly arranged. We model the problem of Procurement Network Formation (PNF) for multiple units of a single item as a cooperative game where agents cooperate to form a surplus maximizing procurement network and then share the surplus in a stable and fair manner. We first investigate the stability of such networks by examining the conditions under which the core of the game is nonempty. We then present a protocol, based on the extensive form game realization of the core, for forming such networks so that the resulting network is stable. We also mention a key result when the Shapley value is applied as a solution concept.

Proceedings of WEBIST 2006 - Second International Conference on Web Information Systems and Technologies, 2006

With major advances in computing technology and network performance, grid computing is strategica... more With major advances in computing technology and network performance, grid computing is strategically placed to become the future of enterprise and even personal computing. One of the most important issues concerned with grid computing is that of application scheduling. The type of scheduling algorithm used will depend on the type of the application. In a global grid setting, the individual users must be provided with an incentive to offer their resources. The situation becomes non-trivial because of the fact that these entities are intelligent, rational and selfish resource providers who, for strategic reasons, may not provide truthful information about their processing power and cost structure. In this scenario, apart from optimality of the algorithm used, strategy-proofness of the underlying mechanism becomes important. This paper presents a strategyproof mechanism based scheduling algorithm for parallel flow type applications in the form of a reverse auction.

2016 International Joint Conference on Neural Networks (IJCNN), Jul 1, 2016

We study the problem of training an accurate linear regression model by procuring labels from mul... more We study the problem of training an accurate linear regression model by procuring labels from multiple noisy crowd annotators, under a budget constraint. We propose a Bayesian model for linear regression in crowdsourcing and use variational inference for parameter estimation. To minimize the number of labels crowdsourced from the annotators, we adopt an active learning approach. In this specific context, we prove the equivalence of well-studied criteria of active learning like entropy minimization and expected error reduction. Interestingly, we observe that we can decouple the problems of identifying an optimal unlabeled instance and identifying an annotator to label it. We observe a useful connection between the multi-armed bandit framework and the annotator selection in active learning. Due to the nature of the distribution of the rewards on the arms, we use the Robust Upper Confidence Bound (UCB) scheme with truncated empirical mean estimator to solve the annotator selection problem. This yields provable guarantees on the regret. We further apply our model to the scenario where annotators are strategic and design suitable incentives to induce them to put in their best efforts.

Lecture Notes in Computer Science, 2004

Combinatorial exchanges are double sided marketplaces with multiple sellers and multiple buyers t... more Combinatorial exchanges are double sided marketplaces with multiple sellers and multiple buyers trading with the help of combinatorial bids. The allocation and other associated problems in such exchanges are known to be among the hardest to solve among all economic mechanisms. In this paper, we study combinatorial exchanges where (1) the demand can be aggregated, for example, a procurement exchange or (2) the supply can be aggregated, for example, an exchange selling excess inventory. We show that the allocation problem in such exchanges can be solved efficiently through decomposition when buyers and sellers are single minded. The proposed approach decomposes the problem into two stages: a forward or a reverse combinatorial auction (stage 1) and an assignment problem (stage 2). The assignment problem in Stage 2 can be solved optimally in polynomial time and thus these exchanges have computational complexity equivalent to that of one sided combinatorial auctions. Through extensive numerical experiments, we show that our approach produces high quality solutions and is computationally efficient.

Sadhana, 2005

Combinatorial auctions (CAs) have recently generated significant interest as an automated mechani... more Combinatorial auctions (CAs) have recently generated significant interest as an automated mechanism for buying and selling bundles of goods. They are proving to be extremely useful in numerous e-business applications such as eselling, e-procurement, e-logistics, and B2B exchanges. In this article, we introduce combinatorial auctions and bring out important issues in the design of combinatorial auctions. We also highlight important contributions in current research in this area. This survey emphasizes combinatorial auctions as applied to electronic business situations.

2008 IEEE International Conference on Automation Science and Engineering, 2008

A business cluster is a co-located group of micro, small, medium scale enterprises. Such firms ca... more A business cluster is a co-located group of micro, small, medium scale enterprises. Such firms can benefit significantly from their co-location through shared infrastructure and shared services. Cost sharing becomes an important issue in such sharing arrangements especially when the firms exhibit strategic behavior. There are many cost sharing methods and mechanisms proposed in the literature based on game theoretic foundations. These mechanisms satisfy a variety of efficiency and fairness properties such as allocative efficiency, budget balance, individual rationality, consumer sovereignty, strategyproofness, and group strategyproofness. In this paper, we motivate the problem of cost sharing in a business cluster with strategic firms and illustrate different cost sharing mechanisms through the example of a cluster of firms sharing a logistics service. Next we look into the problem of a business cluster sharing ICT (information and communication technologies) infrastructure and expl...

Sadhana, 1994

Recently, Brownian networks have emerged as an effective stochastic model to approximate multicIa... more Recently, Brownian networks have emerged as an effective stochastic model to approximate multicIass queueing networks with dynamic scheduling capability, under conditions of balanced heavy loading. This paper is a tutorial introduction to dynamic scheduling in manufacturing systems using Brownian networks.. The article starts with motivational examples. It then provides a review of relevant weak convergence concepts, followed by a description of the limiting behaviour of queueing systems under heavy traffic. The Brownian approximation procedure is discussed in detail and generic case studies are provided to illustrate the procedure and demonstrate its effectiveness. This paper places empha~sis only on the results and aspires to provide the reader with an up-to-date understanding of dynamic scheduling based on Brownian approximations.

Annals of Operations Research, 2006

In this paper, we use reinforcement learning (RL) techniques to determine dynamic prices in an el... more In this paper, we use reinforcement learning (RL) techniques to determine dynamic prices in an electronic monopolistic retail market. The market that we consider consists of two natural segments of customers, captives and shoppers. Captives are mature, loyal buyers whereas the shoppers are more price sensitive and are attracted by sales promotions and volume discounts. The seller is the learning agent in the system and uses RL to learn from the environment. Under (reasonable) assumptions about the arrival process of customers, inventory replenishment policy, and replenishment lead time distribution, the system becomes a Markov decision process thus enabling the use of a wide spectrum of learning algorithms. In this paper, we use the Q-learning algorithm for RL to arrive at optimal dynamic prices that optimize the seller's performance metric (either long term discounted profit or long run average profit per unit time). Our model and methodology can also be used to compute optimal reorder quantity and optimal reorder point for the inventory policy followed by the seller and to compute the optimal volume discounts to be offered to the shoppers.

Lecture Notes in Computer Science, 2005

This paper takes the first steps towards designing incentive compatible mechanisms for hierarchic... more This paper takes the first steps towards designing incentive compatible mechanisms for hierarchical decision making problems involving selfish agents. We call these Stackelberg problems. These are problems where the decisions or actions in successive layers of the hierarchy are taken in a sequential way while decisions or actions within each layer are taken in a simultaneous manner. There are many immediate applications of these problems in distributed computing, grid computing, network routing, ad hoc networks, electronic commerce, and distributed artificial intelligence. We consider a special class of Stackelberg problems called SLRF (Single Leader Rest Followers) problems and investigate the design of incentive compatible mechanisms for these problems. In developing our approach, we are guided by the classical theory of mechanism design. To illustrate the design of incentive compatible mechanisms for Stackelberg problems, we consider first-price and second-price electronic procurement auctions with reserve prices. Using the proposed framework, we derive some interesting results regarding incentive compatibility of these two mechanisms.

Review of Economic Design, 2015

We consider an infinite horizon dynamic mechanism design problem with interdependent valuations. ... more We consider an infinite horizon dynamic mechanism design problem with interdependent valuations. In this setting the type of each agent is assumed to be evolving according to a first order Markov process and is independent of the types of other agents. However, the valuation of an agent can depend on the types of other agents, which makes the problem fall into an interdependent valuation setting. Designing truthful mechanisms in this setting is non-trivial in view of an impossibility result which says that for interdependent valuations, any efficient and ex-post incentive compatible mechanism must be a constant mechanism, even in a static setting. Mezzetti (2004) circumvents this problem by splitting the decisions of allocation and payment into two stages. However, Mezzetti's result is limited to a static setting and moreover in the second stage of that mechanism, agents are weakly indifferent about reporting their valuations truthfully. This paper provides a first attempt at designing a dynamic mechanism which is efficient, strict ex-post incentive compatible and ex-post individually rational in a setting with interdependent values and Markovian type evolution.

arXiv preprint arXiv:1202.3751, Feb 14, 2012

Abstract: The assignment of tasks to multiple resources becomes an interesting game theoretic pro... more Abstract: The assignment of tasks to multiple resources becomes an interesting game theoretic problem, when both the task owner and the resources are strategic. In the classical, nonstrategic setting, where the states of the tasks and resources are observable by the controller, this problem is that of finding an optimal policy for a Markov decision process (MDP). When the states are held by strategic agents, the problem of an efficient task allocation extends beyond that of solving an MDP and becomes that of designing a ...

Consider a requester who wishes to crowdsource a series of identical binary labeling tasks to a p... more Consider a requester who wishes to crowdsource a series of identical binary labeling tasks to a pool of workers so as to achieve an assured accuracy for each task, in a cost optimal way. The workers are heterogeneous with unknown but fixed qualities and their costs are private. The problem is to select for each task an optimal subset of workers so that the outcome obtained after aggregating the labels from the selected workers guarantees a target accuracy level. The problem is a challenging one even in a non strategic setting since the accuracy of aggregated label depends on unknown qualities. We develop a novel multi-armed bandit (MAB) mechanism for solving this problem. First, we propose a framework, Assured Accuracy Bandit (AAB), which leads to a MAB algorithm, Constrained Confidence Bound for a Non Strategic setting (CCB-NS). We derive an upper bound on the number of time steps the algorithm chooses a suboptimal set that depends on the target accuracy level and true qualities. A more challenging situation arises when the requester not only has to learn the qualities of the workers but also elicit their true costs. We modify the CCB-NS algorithm to obtain an adaptive exploration separated algorithm which we call Constrained Confidence Bound for a Strategic setting (CCB-S). CCB-S algorithm produces an ex-post monotone allocation rule and thus can be transformed into an ex-post incentive compatible and ex-post individually rational mechanism that learns the qualities of the workers and guarantees a given target accuracy level in a cost optimal way. We also provide a lower bound on the number of times any algorithm should select a sub-optimal set and we see that the lower bound matches our upper bound upto a constant factor. We provide insights on the practical implementation of this framework through an illustrative example and we show the efficacy of our algorithms through simulations.

Sadhana, 2005

Dynamic pricing is the dynamic adjustment of prices to consumers depending upon the value these c... more Dynamic pricing is the dynamic adjustment of prices to consumers depending upon the value these customers attribute to a product or service. Today's digital economy is ready for dynamic pricing; however recent research has shown that the prices will have to be adjusted in fairly sophisticated ways, based on sound mathematical models, to derive the benefits of dynamic pricing. This article attempts to survey different models that have been used in dynamic pricing. We first motivate dynamic pricing and present underlying concepts, with several examples, and explain conditions under which dynamic pricing is likely to succeed. We then bring out the role of models in computing dynamic prices. The models surveyed include inventory-based models, data-driven models, auctions, and machine learning. We present a detailed example of an e-business market to show the use of reinforcement learning in dynamic pricing.

Annals of Mathematics and Artificial Intelligence, 2019

We study the problem of a buyer (aka auctioneer) who gains stochastic rewards by procuring multip... more We study the problem of a buyer (aka auctioneer) who gains stochastic rewards by procuring multiple units of a service or item from a pool of heterogeneous strategic agents. The reward obtained for a single unit from an allocated agent depends on the inherent quality of the agent; the agent's quality is fixed but unknown. Each agent can only supply a limited number of units (capacity of the agent). The costs incurred per unit and capacities are private information of the agents. The auctioneer is required to elicit costs as well as capacities (making the mechanism design bidimensional) and further, learn the qualities of the agents as well, with a view to maximize her utility. Motivated by this, we design a bidimensional multi-armed bandit procurement auction that seeks to maximize the expected utility of the auctioneer subject to incentive compatibility and individual rationality while simultaneously learning the unknown qualities of the agents. We first assume that the qualities are known and propose an optimal, truthful mechanism 2D-OPT for the auctioneer to elicit costs and capacities. Next, in order to learn the qualities of the agents in addition, we provide sufficient conditions for a learning algorithm to be Bayesian incentive compatible and individually rational. We finally design a novel learning mechanism, 2D-UCB that is stochastic Bayesian incentive compatible and individually rational.

Proceedings 2000 ICRA. Millennium Conference. IEEE International Conference on Robotics and Automation. Symposia Proceedings (Cat. No.00CH37065)

A supply chain network can be viewed as a network of facilities in which a customer order will fl... more A supply chain network can be viewed as a network of facilities in which a customer order will flow through internal business processes such as procurement, production, and transportation, ultimately reaching the required products on time to customers. The delivery performance of such a network can be maximized by pushing the work through the system in a way that the finished products reach the customer in a customer specified delivery window, with a very high probability. This entails synchronization and hence strict control of variability of deliveries at all intermediate points, to ensure that the m w materials and semi-finished work arrive at work spots in a timely fashion. In this paper, we explore the use of the Motorola six sigma tolemncang methodology to achieve synchronization an supply chains. In particular, we use the six sigma approach to: (1) analyze a given supply chain process f o r six sigma delivery performance; and (2) design synchronized supply chains to guarantee six sigma delivery performance. W e provide an example of a plastics industry supply chain, for which we report analysis and design experiments that demonstrate the use of the six sigma approach in designing synchronized supply chains with high levels of delivery performance.

Proceedings of the 19th ACM international conference on Information and knowledge management - CIKM '10, 2010

The advent of large scale online social networks has resulted in a spurt of studies on the user p... more The advent of large scale online social networks has resulted in a spurt of studies on the user participation in the networks. We consider a query incentive model on social networks, where user's queries are answered through her friendship network and there are `rewards' or `incentives' in the system to answer the queries utilizing ones community. We model the friendship network as a random graph with power-law degree distribution, and show that the reward function exhibits a theoretic threshold behavior on the scaling exponent α, a network parameter. Specifically, there exists a threshold on α above which the reward is exponential in the average path length in the network and below the threshold, the reward is proportional to the average path length. We demonstrate this finding on simulated power-law networks and real world data gathered from online social media such as Flickr, Digg, YouTube and Yahoo! Answers.

International Series in Operations Research & Management Science, 2011

In this chapter, we are interested in a procurement network formation problem. We present a case ... more In this chapter, we are interested in a procurement network formation problem. We present a case for modelling the procurement network formation problem as a shortest path cooperative game. We investigate recent results in shortest path cooperative games and their implications to the procurement network formation problem. We then enhance the model for procurement network formation by incorporating asymmetry in the information that agents have. Specifically, we model the procurement network formation problem as a shortest path cooperative game with incomplete information. We point out the incentive compatible core as an appropriate solution concept for this category of games. We then review the current state of the art on the topic of incentive compatible core, pose a conjecture and end with some directions for future work.

Proceedings 2002 IEEE International Conference on Robotics and Automation (Cat. No.02CH37292)

Procurement, i.e., the process of obtaining materials or services and managing their inflow into ... more Procurement, i.e., the process of obtaining materials or services and managing their inflow into organizations, is a critical process in supply chain management. Internet technologies offer attractive opportunities to reduce purchasing costs while increasing the speed and efficiency of the procurement process through seamless network integration and automated execution of key tasks. This paper attempts to model e-procurement from a lead time perspective, to capture the effects of Internet technologies on the procurement lead time. The model we develop is a multiclass queueing network called a probabilistic reentrant line. The model captures several important facets of an eprocurement process, such as use of electronic RFQ (request for quotation) in systematic sourcing, use of Internet auctions in spot purchasing, contention for procurement resources, randomness in execution times of procurement tasks, and concurrent execution of multiple procurement activities. The analysis of the queueing model suggests that end-to-end lead times in eprocurement can be compressed in ingenious ways using standard queueing theoretic techniques.

Proceedings 2003 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS 2003) (Cat. No.03CH37453)

Ahfmcl-In this paper, we come up with an innovative approach through which variability reduction ... more Ahfmcl-In this paper, we come up with an innovative approach through which variability reduction and synchro-Lead time compression in supply chains is the subject of several recent vaoers, see for examde, Gare, Narahari,. .. nimtion can he realized in supply chains. The approach and nswanaham [I], Statistical design toler&ing is a developed is founded on a connection between mechanical design tolerancing and supply chain lead time compmsion, We use two metrics for delivew oerfnrmance. deliverv sham-mature subject in the design community. The key ideas in statistical design tolerancing which provide the core

Advances in Knowledge Discovery and Data Mining, 2010

Game theory is replete with brilliant solution concepts such as the Nash equilibrium, the core, t... more Game theory is replete with brilliant solution concepts such as the Nash equilibrium, the core, the Shapley value, etc. These solution concepts and their extensions are finding widespread use in solving several fundamental problems in knowledge discovery and data mining. The problems include clustering, classification, discovering influential nodes, social network analysis, etc. The first part of the talk will present the conceptual underpinnings underlying the use of game theoretic techniques in such problem solving. The second part of the talk will delve into two problems where we have recently obtained some interesting results: (a) Discovering influential nodes in social networks using the Shapley value, and (b) Identifying topologies of strategically formed social networks using a game theoretic approach.

The 9th IEEE International Conference on E-Commerce Technology and The 4th IEEE International Conference on Enterprise Computing, E-Commerce and E-Services (CEC-EEE 2007), 2007

Formation of high value procurement networks involves a bottom-up assembly of complex production,... more Formation of high value procurement networks involves a bottom-up assembly of complex production, assembly, and exchange relationships through supplier selection and contracting decisions, where suppliers are intelligent and rational agents who act strategically. In this paper we address the problem of forming procurement networks for items with value adding stages that are linearly arranged. We model the problem of Procurement Network Formation (PNF) for multiple units of a single item as a cooperative game where agents cooperate to form a surplus maximizing procurement network and then share the surplus in a stable and fair manner. We first investigate the stability of such networks by examining the conditions under which the core of the game is nonempty. We then present a protocol, based on the extensive form game realization of the core, for forming such networks so that the resulting network is stable. We also mention a key result when the Shapley value is applied as a solution concept.

Proceedings of WEBIST 2006 - Second International Conference on Web Information Systems and Technologies, 2006

With major advances in computing technology and network performance, grid computing is strategica... more With major advances in computing technology and network performance, grid computing is strategically placed to become the future of enterprise and even personal computing. One of the most important issues concerned with grid computing is that of application scheduling. The type of scheduling algorithm used will depend on the type of the application. In a global grid setting, the individual users must be provided with an incentive to offer their resources. The situation becomes non-trivial because of the fact that these entities are intelligent, rational and selfish resource providers who, for strategic reasons, may not provide truthful information about their processing power and cost structure. In this scenario, apart from optimality of the algorithm used, strategy-proofness of the underlying mechanism becomes important. This paper presents a strategyproof mechanism based scheduling algorithm for parallel flow type applications in the form of a reverse auction.

2016 International Joint Conference on Neural Networks (IJCNN), Jul 1, 2016

We study the problem of training an accurate linear regression model by procuring labels from mul... more We study the problem of training an accurate linear regression model by procuring labels from multiple noisy crowd annotators, under a budget constraint. We propose a Bayesian model for linear regression in crowdsourcing and use variational inference for parameter estimation. To minimize the number of labels crowdsourced from the annotators, we adopt an active learning approach. In this specific context, we prove the equivalence of well-studied criteria of active learning like entropy minimization and expected error reduction. Interestingly, we observe that we can decouple the problems of identifying an optimal unlabeled instance and identifying an annotator to label it. We observe a useful connection between the multi-armed bandit framework and the annotator selection in active learning. Due to the nature of the distribution of the rewards on the arms, we use the Robust Upper Confidence Bound (UCB) scheme with truncated empirical mean estimator to solve the annotator selection problem. This yields provable guarantees on the regret. We further apply our model to the scenario where annotators are strategic and design suitable incentives to induce them to put in their best efforts.