Decision-theoretic Techniques Research Articles

Computing value of spatiotemporal information

Location data from mobile devices is a sensitive yet valuable commodity for location-based services and advertising. We investigate the intrinsic value of location data in the context of strong privacy, where location information is only available from end users via purchase. We present an algorithm to compute the expected value of location data from a user, without access to the specific coordinates of the location data point. We use decision-theoretic techniques to provide a principled way for a potential buyer to make purchasing decisions about private user location data. We illustrate our approach in three scenarios: the delivery of targeted ads specific to a user's home location, the estimation of traffic speed, and the prediction of location. In all three cases, the methodology leads to quantifiably better purchasing decisions than competing approaches.

To Buy or Not to Buy

Location data from mobile devices is a sensitive yet valuable commodity for location-based services and advertising. We investigate the intrinsic value of location data in the context of strong privacy, where location information is only available from end users via purchase. We present an algorithm to compute the expected value of location data from a user, without access to the specific coordinates of the location data point. We use decision-theoretic techniques to provide a principled way for a potential buyer to make purchasing decisions about private user location data. We illustrate our approach in three scenarios: the delivery of targeted ads specific to a user’s home location, the estimation of traffic speed, and location prediction. In all three cases, the methodology leads to quantifiably better purchasing decisions than competing methods.

Accounting for asymmetry between strengths, weaknesses, opportunities and threats in outranking methods

PurposeThe purpose of this paper consists in discussing and modeling the asymmetrical nature of evaluation criteria that are notably relevant in strategic planning, project portfolio management and complex systems management.Design/methodology/approachThe author guarantees the theoretic validity of the proposed approach, by showing that it possesses desirable properties such as monotonicity. This allows for a reduced implementation time and thus a real-time use of the model.FindingsThe author conducts a comparative study with two other multicriteria decision aid models on a set of real data regarding a power plant sitting problem and highlights some of the limitations of previous models compared to the proposed approach.Originality/valueThis paper describes a novel multicriteria decision aid model, entitled SPADE (which is an acronym for Structuring Preferences to Aid DEcisions). It aims at consolidating SWOT analysis with sound decision-theoretic techniques, and making use of its results to support decision-making.

Human–agent collaboration for disaster response

In the aftermath of major disasters, first responders are typically overwhelmed with large numbers of, spatially distributed, search and rescue tasks, each with their own requirements. Moreover, responders have to operate in highly uncertain and dynamic environments where new tasks may appear and hazards may be spreading across the disaster space. Hence, rescue missions may need to be re-planned as new information comes in, tasks are completed, or new hazards are discovered. Finding an optimal allocation of resources to complete all the tasks is a major computational challenge. In this paper, we use decision theoretic techniques to solve the task allocation problem posed by emergency response planning and then deploy our solution as part of an agent-based planning tool in real-world field trials. By so doing, we are able to study the interactional issues that arise when humans are guided by an agent. Specifically, we develop an algorithm, based on a multi-agent Markov decision process representation of the task allocation problem and show that it outperforms standard baseline solutions. We then integrate the algorithm into a planning agent that responds to requests for tasks from participants in a mixed-reality location-based game, called AtomicOrchid, that simulates disaster response settings in the real-world. We then run a number of trials of our planning agent and compare it against a purely human driven system. Our analysis of these trials show that human commanders adapt to the planning agent by taking on a more supervisory role and that, by providing humans with the flexibility of requesting plans from the agent, allows them to perform more tasks more efficiently than using purely human interactions to allocate tasks. We also discuss how such flexibility could lead to poor performance if left unchecked.

POMDP-based control of workflows for crowdsourcing

Crowdsourcing, outsourcing of tasks to a crowd of unknown people (“workers”) in an open call, is rapidly rising in popularity. It is already being heavily used by numerous employers (“requesters”) for solving a wide variety of tasks, such as audio transcription, content screening, and labeling training data for machine learning. However, quality control of such tasks continues to be a key challenge because of the high variability in worker quality. In this paper we show the value of decision-theoretic techniques for the problem of optimizing workflows used in crowdsourcing. In particular, we design AI agents that use Bayesian network learning and inference in combination with Partially-Observable Markov Decision Processes (POMDPs) for obtaining excellent cost-quality tradeoffs.We use these techniques for three distinct crowdsourcing scenarios: (1) control of voting to answer a binary-choice question, (2) control of an iterative improvement workflow, and (3) control of switching between alternate workflows for a task. In each scenario, we design a Bayes net model that relates worker competency, task difficulty and worker response quality. We also design a POMDP for each task, whose solution provides the dynamic control policy. We demonstrate the usefulness of our models and agents in live experiments on Amazon Mechanical Turk. We consistently achieve superior quality results than non-adaptive controllers, while incurring equal or less cost.

CAPIR: Collaborative Action Planning with Intention Recognition

We apply decision theoretic techniques to construct non-player characters that are able to assist a human player in collaborative games. The method is based on solving Markov decision processes, which can be difficult when the game state is described by many variables. To scale to more complex games, the method allows decomposition of a game task into subtasks, each of which can be modelled by a Markov decision process. Intention recognition is used to infer the subtask that the human is currently performing, allowing the helper to assist the human in performing the correct task. Experiments show that the method can be effective, giving near-human level performance in helping a human in a collaborative game.

Propositions To Guide Evidence-Based Decision-Making

<p class="MsoNormal" style="text-align: justify; margin: 0in 0.5in 0pt;"><span style="font-size: 10pt;"><span style="font-family: Times New Roman;">The methods by which managers make decisions in the face of inaccurate data, changing environments and other pervasive uncertainties, have been studied by researchers and business practitioners for many decades.<span style="mso-spacerun: yes;">  </span>Much attention is given to highly quantitative decision-theoretic techniques; such methods generally deal with these inherent impediments to rational decisions using mathematical concepts from the theories of probability, stochastic processes, estimation and optimization, fuzzy sets, etc.<span style="mso-spacerun: yes;">  </span>The very act of quantification itself has great implications for managers’ cognitive processes, the impact upon various groups within the organization, and the final outcomes.<span style="mso-spacerun: yes;">  </span>Indeed, use of formal methods and actual reliance on hard numbers improves decision quality and speed by providing efficient cognitive simplifications and convergent expectations.<span style="mso-spacerun: yes;">  </span>This paper synthesizes findings from several relevant streams of literature and proposes several simple propositions for further discussion and future research. It applies these ideas to two illustrative examples of major complex business decisions.</span></span></p>

Design for configurability: rethinking interdomain routing policies from the ground up

Giving ISPs more fine-grain control over interdomain routing policies would help them better manage their networks and offer value-added services to their customers. Unfortunately, the current BGP route-selection process imposes inherent restrictions on the policies an ISP can configure, making many useful policies infeasible. In this paper, we present Morpheus, a routing control platform that is designed for configurability. Morpheus enables a single ISP to safely realize a much broader range of routing policies without requiring changes to the underlying routers or the BGP protocol itself. Morpheus allows network operators to: (1) make flexible trade-offs between policy objectives through a weighted-sum based decision process, (2) realize customer-specific policies by supporting multiple route-selection processes in parallel, and allowing customers to influence the decision processes, and (3) configure the decision processes through a simple and intuitive configuration interface based on the Analytic Hierarchy Process, a decision-theoretic technique for balancing conflicting objectives. We also present the design, implementation, and evaluation of Morpheus as an extension to the XORP software router.

Decisions in thesaurus construction and use

A thesaurus and an ontology provide a set of structured terms, phrases, and metadata, often in a hierarchical arrangement, that may be used to index, search, and mine documents. We describe the decisions that should be made when including a term, deciding whether a term should be subdivided into its subclasses, or determining which of more than one set of possible subclasses should be used. Based on retrospective measurements or estimates of future performance when using thesaurus terms in document ordering, decisions are made so as to maximize performance. These decisions may be used in the automatic construction of a thesaurus. The evaluation of an existing thesaurus is described, consistent with the decision criteria developed here. These kinds of user-focused decision-theoretic techniques may be applied to other hierarchical applications, such as faceted classification systems used in information architecture or the use of hierarchical terms in “breadcrumb navigation”.

Is 1 noun worth 2 adjectives? Measuring relative feature utility

Are two adjectives worth the same as a single noun when documents are ordered based on decreasing topicality? We propose an easy to interpret single number Relative Feature Utility (RFU) measure of the relative worth of using specific linguistic or non-linguistic features or sets of features in computational systems that order or filter media, such as information retrieval and classification systems. This measure allows one to make easily interpreted claims about the relative utility of features such as parts-of-speech, term suffixes, phrases vs. single terms, annotations, hyperlinks, citations, index terms, and metadata when ordering natural language text or other media. Data is provided for the RFU for stemming characteristics, part-of-speech tags, and phrase lengths, as well as retrieval characteristics and procedures. Using this linear measure of the relative utility of features makes available a wide range of cost-benefit analyses and decision theoretic techniques, allowing the study of whether or not to use many different kinds of representational information or tagging systems, and for the design of indexing and metadata systems. Some characteristics of natural languages used in the spectrum from softer to harder sciences, as well as medical terminology, are studied.

Bridging the gap between statistical analysis and decision making in public health research

Statisticians have eagerly taken on the role of presenting statistical summaries of quantitative data. In areas of health, this means providing point and interval estimates for quantities of interest such as diagnostic risks and treatment effects or providing curve estimates for quantities of interest such as survival probabilities across time. Methods for providing such summaries are highly formalized and constantly evolving. While decision making is the incentive for nearly all such efforts, the process that transforms statistical summaries into decisions usually remains informal and ad hoc. Statisticians have not eagerly accepted the role of promoting formalized decision-theoretic techniques. This paper will argue that the gap between statistical synthesis and decision making is an unnatural and undesirable one, because it undermines the impact of quantitative information. An argument for bridging the gap by expanding the role of statisticians will be presented.

Framework of a Decision-Theoretic Tutoring System for Learning of Mechanics.

This paper presents the application of decision-theoretic technique to computer-based tutoring system for elementary mechanics. The technique uses sound probabilistic reasoning and a student model to identify learner's misconception(s). Bayesian belief networks are the building blocks of the student model. The probability values in Bayes' nets are provided by teacher and are based on her judgement, but may be substituted with actual statistics. Evidence on student's mastery of concepts is obtained through her responses to appropriately selected items. Subsequently, Rasch one-parameter model is used to calibrate the item and person parameters (also known as difficulty and ability indices, respectively). The system is able to provide teacher with information for fine-tuning her pedagogical instructions and guide her in coaching students. It is also able to provide students with immediate feedback to improve their proficiencies and ultimately their grades.

Towards Flexible Teamwork in Persistent Teams: Extended Report

Teamwork is a critical capability in multi-agent environments. Many such environments mandate that the agents and agent-teams must be persistent i.e., exist over long periods of time. Agents in such persistent teams are bound together by their long-term common interests and goals. This paper focuses on flexible teamwork in such persistent teams. Unfortunately, while previous work has investigated flexible teamwork, persistent teams remain unexplored. For flexible teamwork, one promising approach that has emerged is model-based, i.e., providing agents with general models of teamwork that explicitly specify their commitments in teamwork. Such models enable agents to autonomously reason about coordination. Unfortunately, for persistent teams, such models may lead to coordination and communication actions that while locally optimal, are highly problematic for the team's long-term goals. We present a decision-theoretic technique based on Markov decision processes to enable persistent teams to overcome such limitations of the model-based approach. In particular, agents reason about expected team utilities of future team states that are projected to result from actions recommended by the teamwork model, as well as lower-cost (or higher-cost) variations on these actions. To accommodate real-time constraints, this reasoning is done in an any-time fashion. Implemented examples from an analytic search tree and some real-world domains are presented.

Integrating case-based reasoning and decision theory

Case-based reasoning (CBR) and decision-theoretic techniques can be complementary. Decision theory helps CBR deal with uncertainties in the problem domain, while CBR helps decision theory handle complicated problems with many variables. The goal of integrating CBR and decision theory is to improve the ability of CBR systems to solve problems in domains of incomplete information. Our methodology views the retrieval of old cases in CBR as a decision problem, where each case from the case base provides an alternative solution and a prediction of the possible outcomes for the problem. When case-based problem solving encounters uncertainty, our methodology applies decision theory to evaluate each case in terms of the attributes that are significant for the problem, so that the most desirable case can be selected. We implemented our methodology in a case-based design assistant that helps chemists design pharmaceuticals.

A probabilistic reasoning model: Formulation and control strategy

It has been recognized that past experiences of a decision maker often plays a pivotal role in solving new problem instances. Therefore, the ability to model human reasoning processes has become an important subject of research in recent years. In many applications, the reasoning process must deal with uncertainty inherent in the problem domain. This research addresses the issue of supporting the model formulation and data acquisition processes for situations that (i) operate under uncertain conditions, and (ii) utilize evidential information that is gathered in stages. A theoretical framework is presented for the probabilistic formulation of the reasoning process that incorporates past experiences. The model is validated by testing its performance on simulated data, and is shown to work well when a sufficiently large number of cases are available for estimating probabilities. The probabilistic reasoning system can revise beliefs in an intuitively appealing and theoretically sound manner when information is acquired in an incremental fashion. Two dynamic information gathering strategies are discussed for such a reasoning system, one using information theoretic techniques, and the other using decision theoretic techniques.

Control of selective perception using bayes nets and decision theory

A selective vision system sequentially collects evidence to support a specified hypothesis about a scene, as long as the additional evidence is worth the effort of obtaining it. Efficiency comes from processing the scene only where necessary, to the level of detail necessary, and with only the necessary operators. Knowledge representation and sequential decision-making are central issues for selective vision, which takes advantage of prior knowledge of a domain''s abstract and geometrical structure and models for the expected performance and cost of visual operators. .pp The TEA-1 selective vision system uses Bayes nets for representation and benefit-cost analysis for control of visual and non-visual actions. It is the high-level control for an active vision system, enabling purposive behavior, the use of qualitative vision modules and a pointable multiresolution sensor. TEA-1 demonstrates that Bayes nets and decision theoretic techniques provide a general, re-usable framework for constructing computer vision systems that are selective perception systems, and that Bayes nets provide a general framework for representing visual tasks. Control, or decision making, is the most important issue in a selective vision system. TEA-1''s decisions about what to do next are based on general hand-crafted ``goodness functions'''' constructed around core decision theoretic elements. Several goodness functions for different decisions are presented and evaluated. .pp The TEA-1 system solves a version of the T-world problem, an abstraction of a large set of domains and tasks. Some key factors that affect the success of selective perception are analyzed by examining how each factor affects the overall performance of TEA-1 when solving ensembles of randomly produced, simulated T-world domains and tasks. TEA-1''s decision making algorithms are also evaluated in this manner. Experiments in the lab for one specific T-world domain, table settings, are also presented.

Machine learning for intelligent support of conflict resolution

Because negotiation and conflict resolution are complex and unstructured tasks, they need sophisticated decision support. One of the crucial characteristics of such support is systems that are capable of improving their performance, both in terms of efficiency and solution quality, by employing machine learning techniques. A framework for intelligent computer-supported conflict resolution through negotiation/ mediation is presented. The model integrates Artificial Intelligence methods (case-based reasoning) and decision theoretic techniques (multi-attribute utilities) to provide enhanced conflict resolution and negotiation support in group problem solving. This model has been implemented in the PERSUADER, a computer program which operates in the domain of resolution of labor management disputes. The PERSUADER uses case-based reasoning (CBR) to learn from its experience. In contrast to quantitative models or expert systems that solve each problem from scratch and discard the solution at the end of problem solving, CBR retains the process and results of its computational decisions so that they can be re-used to solve future related problems. CBR is a powerful learning method since it enables a system not only to exploit previous succesful decisions, thus short-cutting possibly long reasoning chains, but also to profit from previous failures by using them to recognize similar failures in advance so they can be avoided in the future. As the state of the art in DSS development advances and as DSSs support increasingly more complicated tasks, such machine learning techniques will become an indispensable part of decision support systems.

A Decision-Theoretic Technique for the Illumination of Cultural Differences

At a conference of U.S. and Japanese social scientists, a decision-theoretic, multiattribute utilities methodology was used to elucidate values and beliefs of each group. The ten-step process requires each group to specify their values and to rank order and importance weight them. Then a series of judgments are made about the extent to which a list of action alternatives is likely to maximize each value. Compared with the Japanese, the U.S. group was vociferously individualistic. The U.S. group used more extreme importance weights. The Japanese values emphasized the importance of the third world and improved international trade. Both groups judged energy resources, population policies, and economic and arms issues to be highly important. The method was useful in illuminating cultural differences.

An Application of Decision Theory to a Medical Diagnosis-Treatment Problem

The decision faced by a physician when confronted by a patient with an undetermined disease may be simply stated as: What course of action, in the form of diagnostic tests and/or treatments, should be taken? In most cases, this problem can be characterized as a sequential decision under uncertainty. Since this is a class of problems for which decision theory has proved a useful tool, it appears fruitful to attempt to apply it to the physician's problem. In this paper, this possibility is explored by describing the application of decision theoretic techniques to a specific case. We first comment on why we believe the proposed model is more appropriate than other methods of treating the problem. Then the proposed model is briefly described in the abstract. The main body of the paper describes a specific problem and its solution by decision theoretic techniques. In the final section, some of the shortcomings of the particular analysis and some of the problems that might be encountered in a more general setting are pointed out.