Information Density in Decision Analysis

Efforts to ensure that funded research represents "value for money" have led to increasing calls for the use of analytic methods in research prioritization. A number of analytic approaches have been proposed to assist research funding decisions, the most prominent of which are value of information (VOI) and prospective payback of research (PPoR). Despite the increasing interest in the topic, there are insufficient VOI and PPoR applications on the same case study to contrast their methods and compare their outcomes. We undertook VOI and PPoR analyses to determine the value of conducting 2 proposed research programs. The application served as a vehicle for identifying differences and similarities between the methods, provided insight into the assumptions and practical requirements of undertaking prospective analyses for research prioritization, and highlighted areas for future research. VOI and PPoR were applied to case studies representing proposals for clinical trials in advanced non-small-cell lung cancer and prostate cancer. Decision models were built to synthesize the evidence available prior to the funding decision. VOI (expected value of perfect and sample information) and PPoR (PATHS model) analyses were undertaken using the developed models. VOI and PPoR results agreed in direction, suggesting that the proposed trials would be cost-effective investments. However, results differed in magnitude, largely due to the way each method conceptualizes the possible outcomes of further research and the implementation of research results in practice. Compared with VOI, PPoR is less complex but requires more assumptions. Although the approaches are not free from limitations, they can provide useful input for research funding decisions.

Decision makers within a jurisdiction facing evidence of positive but uncertain incremental net benefit of a new health care intervention have viable options where no further evidence is anticipated to: (1) adopt the new intervention without further evidence; (2) adopt the new intervention and undertake a trial; or (3) delay the decision and undertake a trial. Value of information methods have been shown previously to allow optimal design of clinical trials in comparing option (2) against option (1), by trading off the expected value and cost of sample information. However, this previous research has not considered the effect of cost of reversal on expected value of information in comparing these options. This paper demonstrates that, where a new intervention is adopted, the expected value of information is reduced under optimal decision making with costs of reversing decisions. Further, the paper shows that comparing expected net gain of optimally designed trials for option (2) vs (1) conditional on cost of reversal, and (3) vs (1) conditional on opportunity cost of delay allow systematic identification of an optimal decision strategy and trial design. Copyright © 2006 John Wiley & Sons, Ltd.

The cost of obtaining good information regarding the various probability distributions needed for the solution of most stochastic decision problems is considerable. It is important to consider questions such as: (1) what minimal amounts of information are sufficient to determine optimal decision rules; (2) what is the value of obtaining knowledge of the actual realization of the random vectors; and (3) what is the value of obtaining some partial information regarding the actual realization of the random vectors. This paper is primarily concerned with questions two and three when the decision maker has an a priori knowledge of the joint distribution function of the random variables. Some remarks are made regarding results along the lines of question one. Mention is made of assumptions sufficient so that knowledge of means, or of means, variances, co-variances and n-moments are sufficient for the calculation of optimal decision rules. The analysis of the second question leads to the development of bounds on the value of perfect information. For multiperiod problems it is important to consider when the perfect information is available. Jensen's inequality is the key tool of the analysis. The calculation of the bounds requires the solution of nonlinear programs and the numerical evaluation of certain functions. Generally speaking, tighter bounds may be obtained only at the expense of additional information and computational complexity. Hence, one may wish to compute some simple bounds to decide upon the advisability of obtaining more information. For the analysis of the value of partial information it is convenient to introduce the notion of a signal. Each signal represents the receipt of certain information, and these signals are drawn from a given probability distribution. When a signal is received, it alters the decision maker's perception of the probability distributions inherent in his decision problem. The choice between different information structures must then take into account these probability distributions as well as the decision maker's preference function. A hierarchy of bounds may be determined for partial information evaluation utilizing the tools of the multiperiod perfect information case. However, the calculation of these bounds is generally considerably more dicult than the calculation of similar boulids in the perfect information case. Most of the analysis is directed towards problems in which the decision maker has a linear utility function over profits, costs or some other numerical variable. However, some of the bounds generalize to the case when the utility function is strictly increasing and concave.

This article considers an agricultural production model of sequential nitrogen application under risk. Because of random shocks between subsequent production stages, optimal fertilization decisions depend on the magnitude of farmers' risk aversion (risk premium), and the possibility for farmers to process information (value of information). We propose a joint estimation procedure of technology and risk aversion parameters, using a structural, simulation‐based econometric procedure. Parameter estimates allow to compute both the value of information and the risk premium, which account for about 30% of fertilizer cost for Midwest corn producers.

Probabilistic sensitivity measures as information value

The research community generally recognizes the Value of Information (VoI) as a realistic indicator of the benefit of the information provided by a monitoring system in structures management. Even though many authors have recently focused on the assessment of the VoI, most of their works do not consider the presence of two individuals in the decision process of whether to install a monitoring system, the owner and the manager of the structure. In this contribution, we define a framework to quantifying the VoI of a monitoring system by considering these two separate individuals, who have a different aversion toward risk. Moreover, we apply the results on a real-life case study concerning the Streicker Bridge in Princeton, NJ, in the USA. This framework aims to help the owner in quantifying the money saved by entrusting the evaluation of the state of the structure to the monitoring system, even if the manager’s behaviour toward risk is different from the owner’s own, and so are his or her management decisions. The results of the case study confirm the difference in the two way to quantify the VoI of a monitoring system.

Discretization of time series and encoding it as a string in a finite alphabet allows application of information theory methods developed for discrete signals. Computing information values of n-grams extracted from such string leads to introduction of events as occurrences of n-grams that possess specific properties, e.g. abnormally high (or low) information value. We define information value of an n-gram via maximum entropy lifts over frequency dictionaries. We also look for correlation between market events and n-gram events. The paper shows that the proposed method of time series analysis when applied to events study may provide new insightful perspective.

Value of information is a widely accepted methodology for evaluating the need to acquire new data in the oil and gas industry. In the conventional approach to estimating the value of information, the outcomes of a project assessment relate to the decision reached by following Boolean logic. However, human thinking is based on a more complex logic that includes the ability to process uncertainty. In value of information assessment, it is often desirable to make decisions based on multiple economic criteria which, if independently evaluated, may suggest opposite decisions. Artificial intelligence has been used successfully in several areas of knowledge, increasing and enhancing analytical capabilities. This paper aims at enriching the value of information methodology by integrating fuzzy logic into the decision-making process; this integration makes it possible to develop a human thinking assessment and coherently combine several economic criteria. To the authors’ knowledge, this is the first use of a fuzzy inference system in the specified knowledge domain. The methodology is successfully applied to a case study of an oil and gas subsurface assessment where the results of the standard and fuzzy methodologies are compared, leading to a more robust and complete evaluation. Sensitivity analysis is undertaken for several membership functions used in the case study to assess the impact that shifting, narrowing and stretching the membership relationship has on the value of information. The results of the sensitivity study show that, depending on the shifting, the membership functions lead to different decisions; additional sensitivities to the type of membership functions are investigated, including the functions’ parameters.

Geoengineering prognoses are often based on data from a limited number of investigations of soil and rock mass. There is generally a desire to reduce the uncertainty in the prognoses while minimising the investigation costs. Value of Information Analysis (VOIA) is a support for decisions regarding investigation strategies and the aim of this paper is to present methodology for VOIA that takes into account four decision alternatives where the input data could be provided by experts. The methodology will be applied in a case study where the value of information related to an investigation borehole will be calculated. The results indicate that the value of information of the borehole is low compared with the realisation costs of the investigation. It was found that models for VOIA in underground construction projects are complex but that the analysis can be simplified with extensive use of expert knowledge and calculations of the value of perfect information as a benchmark for investigation strategies.

Quantifying the value of structural health monitoring information with measurement bias impacts in the framework of dynamic Bayesian Network

At any stage of the Exploration and Production cycle decisions must be made with limited information. This situation can always be improved by investing more money or time in improving the quality and variety of information available. However, this in itself presents a decision as to whether the investment to get new information such as a new seismic survey is worth it. This paper outlines a practical and structured method to estimate the value of new information in monetary terms, which can be used to make an informed business decision on whether the new information is worth it.The method is comprised of two primary elements. Firstly the problem is defined using a decision tree analysis approach that considers what decisions are to be made and their possible outcomes. Secondly, a collaborative workshop is used to estimate the probabilities required to populate the decision tree. This then enables the Expected Monetary Value to be calculated both with and without the technology that acquires the new information. The difference between the two is approximated as the value of the information. This paper uses a case study approach to explain this method that considers an operators dilemma in deciding whether or not to drill an appraisal well.Obtaining a meaningful result from this process requires time and commitment from management and technical staff. However, at stake are the considerable sums of money in acquiring the new information and the subsequent investment decisions that draw on the new information. The method described as a minimum provides fiscal information to make investment decisions and at best, provides a framework in which the benefits and risks of investing in new technology can be fully understood.There has already been much published on the use of decision trees to estimate the value of information. However, in practice it can be difficult to estimate the probabilities required for a decision tree analysis to work as they are not intuitive to describing the situation. This approach simplifies this by arranging the mathematics to make the task of the workshops clearer and more achievable. The result is a practical method that can be easily understood and implemented.

In sequential field development planning, past decisions not only directly affect the maximum achievable expected NPV but also influence the future information that can be used to reduce geological uncertainty. To act optimally, when choosing actions, we must also take into account the opportunities to improve the optimal strategy by reducing future uncertainty. In most applications, however, the effect of future information on the optimal decisions is ignored because it would be computationally intractable to update the reservoir model and re-optimize to account for all possible outcomes of future observations. To efficiently make optimal decisions while considering future possibilities for learning through actions, we developed a flexible workflow built on the key-feature-based value of information (VOI) analysis, which is obtained by identifying key reservoir features for optimization problems and key observations for improving future decisions. Instead of considering future information from all remaining actions, we only consider the important information from key actions to reduce the uncertainty with the largest influence on the optimal strategy – that which would be most helpful in improving future decisions. The efficiency of the method results from the focus on the use of key observations to reduce key uncertainty, rather than using all observations to reduce all uncertainties. In this work, we built supervised-learning algorithms to identify the optimal combination of observations for reducing key uncertainty and simultaneously to estimate the information's reliability. This allows automatic detection of key observations and direct computation of the posterior probability distribution of key uncertainty based on Bayes' rule, avoiding the need for full history matching to re-estimate the uncertainty. Moreover, the entire key observation space is divided into a limited number of disjointed subspaces, such that observations located in the same subspace have almost the same prediction precision for key uncertainty reduction. It is then only necessary to update the reservoir model for each subspace instead of for all distinct sets of observations. Our methods are illustrated by the application of the drilling-order problem in a synthetic field model, for which the drilling sequence of wells is an important contributor to the reservoir's profitability and for which the optimal solution changes significantly with key reservoir features. Results show that using such a simplified VOI analysis based on key actions and key observations can efficiently improve the expected outcome of an optimal strategy with very little performance loss. Although the key actions provide important information for key uncertainty reduction, taking key action rather than the initial optimal decision for the current uncertainty state is not always worthwhile even if the information is obtained without explicit cost. Since there may be an indirect cost of information caused by taking an action that appears to be sub-optimal based on past information, it is necessary to consider both the possibility of key uncertainty reduction and the possibility of high expected NPV to determine whether it is worth taking the action to improve future decisions.

With the rapid development of information technology, industrial Internet has also entered a rapid development stage. As cloud servers have huge storage space as well as powerful computing power, more and more resource-constrained factories outsource their data to cloud servers for computation and storage to reduce the burden of data owners. Industrial data contains information about industrially produced products, and the high value of such information attracts attacks from illegal users, which may lead to data leakage. In addition, cloud servers are not fully trustworthy; therefore, it is important to ensure the security and correctness of industrial Internet data during outsourced computation and storage for the healthy development of the industrial Internet. In this paper, we explore technical solutions to guarantee the security and correctness of industrial Internet data when outsourcing computation, to achieve the privacy protection of industrial Internet data.

My task is to provide a perspective regarding the general topic of choosing optimal cost drivers, as that topic is reflected in the selected articles of this Forum. This perspective can be developed more easily if I first digress back to the 1960s when managerial accounting research was beginning to evolve to its present stages. In the 1960s, teaching and research in managerial accounting were heavily influenced by the growing popularity of operations research and management science approaches to business problems. These approaches were largely normative in nature, consisting basically of optimization techniques that, if implemented, promised to improve managerial decision making. Their normative character was carried over into managerial accounting as well, with the basic theme of many managerial accounting articles in that era being, "Here is how optimization techniques may be used to guide the design of improved managerial accounting systems." Rarely, if ever, did the researcher provide empirical evidence or other forms of proof that the new systems would in fact lead to improvements in managerial decisions. Rather, the main result provided was that the new systems could generate new accounting numbers under certain assumed conditions. If these new numbers appeared to be materially different from the old, implementation of the new systems could lead to changes in managers' decisions. The changes in managers' decisions were assumed to be desirable ones to induce. However, desirability was based primarily on subjective criteria adopted by the researcher. Generally, a necessary condition for potential improvements in managerial decisions from new accounting systems is that the new system generate accounting numbers that are materially different from those obtained from the existing system. For this reason, I label such studies as materiality studies, since they basically follow the traditional materiality rule in accounting. But different accounting numbers might not lead to changes in decisions that result in payoffs sufficient to justify the costs of implementing the new systems. Indeed, recognition of this basic limitation of materiality studies was a strong motivation for managerial accounting researchers to investigate the advantages of using the information economics paradigm to estimate the net payoffs from implementing new accounting systems. Recall that information economics is a conceptual framework that can be used to assess how new information will affect managerial decisions and, assuming decisions will change, the expected payoffs from implementing a new accounting information system. Included in the framework Is an assumed decision-maker with a particular utility function who processes information through a specific decision model. The model, in turn, yields an ex ante optimal decision. The value of new information is then measured as the change in the net payoffs induced by moving from a previous optimal decision to a revised optimal decision under the assumption that the new information does in fact induce a revision in the optimal decision. Note that the original optimal decision might still remain optimal even though the new information received by the decision maker differs from the old. Obviously, the conclusion in the latter case is that the new information system would have either a zero or negative expected net value.

Value of information techniques form an alternative framework for study design, based on the uncertainty in long term costs and benefits associated with different treatment approaches. Given a set of candidate study designs, the optimal design is that which maximises the expected net benefit of sampling (ENBS). Much contemporary discussion has focussed on the potential for using value of information in designing randomised clinical trials for drug development. However, this approach could theoretically be used to design research to improve clinical practice after treatment strategies have been adopted, where long term costs and benefits are still subject to uncertainty. Surgery is recognised as being vital in the treatment of early breast cancer. However, in the UK many elderly women are instead treated with primary endocrine therapy. According to NICE guidance, this constitutes undertreatment. For women with limited life expectancy or frailty, less intensive therapeutic approaches may be beneficial. However, more research is required to ensure that treatments are chosen appropriately. This poster presents a case study of using value of information methods to design a study to address this problem. The use of Bayesian evidence synthesis techniques and simulation allow for the calculation of ENBS of a wide range of study designs, including experimental and observational designs. A key methodological challenge in this case study is specifying a sensible representation of current beliefs about heterogeneity in outcomes given limited evidence. The computational cost of these methods remains a barrier to implementation, though more efficient algorithms are in development.