Applied Decision Support with Soft Computing

Distributed decision support and organizational connectivity: A case study

Decision Support Systems (DSS) in health care must be designed in consideration of the empirical context and problem space where they are being applied. However differences such as workflow, clinical expertise and organizational norms make it hard to define the context where a DSS will be used. What is needed is a DSS that is able to enhance health care delivery in different contexts. In this paper we present a multi-modal DSS that supports decision making in different contexts. The paper describes the theoretical basis for the DSS, explores a problem space in palliative care and describes a prototype implementation of the DSS to address the issues from the problem space.

This paper aims to introduce the Decision Support System (DSS), which is an area informatics approach to area studies, and the research network complex based on DSS's application in Thailand. In particular, the paper presents and discusses current activities of the DSS research network at the national level in Thailand rather than being a theoretical and analytical study of the DSS mechanism or its cases. DSS has wide applications in Thailand, extending to various fields such as agricultural and natural resource management, economic and social management, historical and cultural preservation, and so on. DSS generally refers to a support system embedded into a computer system for providing intellectual resources with the appropriate numerical model necessary for decision making in production activities. Agricultural and natural resources have been the foundation of social and economic development in Thailand since the country's first national plan in the1960s. Decision making to maintain agricultural productivity as well as to protect natural resources requires well-integrated data sets. The Thailand Research Fund(TRF) established a DSS research and development network (TRF-DSS) in 2002 to help various research teams in Thailand develop DSS tools and components for addressing agricultural and natural resource management issues. Using a "systems"approach, the DSS framework allows researchers and users to identify and integrate key components as well as to define databases and model-base management systems. This paper focuses on the TRF-DSS research and development network, which consists of 12 universities, two line agencies in Thailand, and a line agency in Cambodia. During 2002–10, a total of 59 research projects were granted a budget allocation of 140.1 million baht, and a budget of 15.4 million baht was allocated to support activities of the network. In addition, 10 projects were funded to carry out post-project activities, with a total budget of 1.5 million baht. More than 20 DSS tools were developed and implemented by various users, ranging from policy makers to provincial and local government agencies engaged in short- and long-term planning and management. Most DSS tools were designed to allow the integration of biophysical and socioeconomic data as well as the decision support modules for alternatives evaluation and analysis. These approaches support the choices of dynamic simulation models as well as multi-criteria analyses for modelbase software development. They also allow users to evaluate various alternatives in agricultural and natural resource management. Networking is a powerful platform for DSS research and development and maybe applied to other types of research and development agenda in Thailand, such as area study projects. DSS tools contribute to an understanding of sustainability of production systems against a background of climate change, poverty reduction, and food security.

This is the first book to develop a decision support methodology for strategic environmental decision problems, and provides several generic as well as specific tools. The book is divided into three parts plus an Appendix. Part I introduces the methodological background and describes various features of the decision environment and the ways in which model-based decision support can help the decision making process. The methodology of building and analyzing mathematical models that represent underlying physical and economic processes, and that are useful for modern decision makers at various stages of decision making, is presented. These methods support not only the analysis of compromise solutions that correspond best to decision maker preferences but also allow the use of other modeling concepts like soft constraints, soft simulation, or inverse simulation. Part II describes various types of tools that are used for the development of decision support systems. Several of these tools described in the Appendix are available from IIASA's Web site, free of charge, for use in research and teaching. The last part of the book consists of four chapters on applications. Each chapter treats an area of environmental decision making: water quality management in river basins, land use planning, cost effective policies for improving air quality, and energy planning. For each area decision support systems are presented and it is shown how they are used for supporting decision making and negotiations. The applications as well as the methodology presented in this book have been developed at IIASA in close cooperation with several other institutes and organizations.

In this paper we make an extensive study of Artificial Intelligence (AI) techniques that can be used in decision support systems in healthcare. In particular, we propose variants of ensemble methods (i.e., Rotation Forest and Input Decimated Ensembles) that are based on perturbing features, and we make a wide comparison among the ensemble approaches. We illustrate the power of these techniques by applying our approaches to different healthcare problems. Included in this chapter is extensive background material on the single classifier systems, ensemble methods, and feature transforms used in the experimental section.

Changing systems for supporting farmers' decisions: problems, paradigms, and prospects

Integrated pest management (IPM) prioritizes non-chemical alternatives for managing pests and/or diseases, reserving chemicals as last solution, when economically and environmentally justified. The holistic consideration of the agroecosystem makes decision-making in crop protection more complex, as it requires selecting the most appropriate action at the right time. Decision support systems (DSS) for IPM help manage the risks of pests and diseases affecting crops by predicting potential threats and suggesting suitable control measures. However, their adoption remains modest among farmers, despite demonstrated benefits. An anonymous online survey was conducted among farmers, advisors, and other agricultural stakeholders to investigate factors influencing DSS uptake, or lack thereof. The survey distinguished between DSS users and non-users, enabling targeted questions on different aspects of DSS use and adoption. In total, 78 respondents were classified as DSS users and 29 as DSS non-users . Data were analyzed through clustering, followed by statistical validation of clusters and induction of a classification tree, to identify the most influential features shaping cluster formation among both DSS users and DSS non-users . The analysis identified four clusters in the DSS users’ dataset and two clusters in the DSS non-users’ dataset. For DSS users , DSS output reliability and usefulness were identified as the most influential features. Membership in a farmers’ network and DSS ease of use also contributed to cluster separation. For DSS non-users , DSS availability for different cropping systems was the most important attribute determining cluster separation. The results indicate that farmer involvement in DSS development and calibration may enhance trust and, therefore, adoption. Moreover, the need for user-friendly systems providing simple and clear information was highlighted. Although the lack of DSS tailored to specific cropping systems emerged as the main barrier, limited awareness and concerns about costs were also important obstacles. The applied methodological approach presents a reproducible and robust AI-based framework, built on open-source tools, to analyze stakeholders’ perceptions of IPM DSS and extract knowledge from complex survey data. This study provides valuable insights to facilitate DSS adoption for more sustainable crop protection and underscores the importance of both DSS technical performance and end-user needs.

Product development decisions represent challenging processes with various stakeholders from different disciplines and with different perspectives involved. Decisions can be highly complex due to the high amount of influencing factors that have to be considered simultaneously. Among other methods and tools, decision support systems (DSS) were developed to support decision-makers in handling complex decisions. One main disadvantage of DSS is the little adaptation to and integration of decision-maker needs and characteristics. Thus, current trends focus on the development of human- and decision-maker-centered DSS. This paper aims in providing a multilevel stepwise procedure for the modeling of decision-makers in order to enable a more effective decision support. The decision-maker modeling procedure is based on findings from literature research and results from research projects performed in cooperation with industry partners. The procedure serves as a guideline for generating diverse decision-maker models for different applications and for integrating human cognitive thinking procedures within the design and functionality of decision support methodologies and systems. The procedure represents an important design element of the human-centered decision support methodology currently being developed at the Institute of Product Development of the Technische Universität München (Germany).

A SDSS combines database storage technologies, geographic information systems (GIS) and decision modeling into tools which can be used to address a wide variety of decision support areas (Eklund, Kirkby, and Pollitt, 1996). Recently, various emerging technologies in computer hardware and software such as speedy microprocessors, gigabit network connections, fast internet mapping servers along with Web-based technologies like extensible markup language (XML), Web services, etc provide promising opportunities to take the traditional spatial decision support systems one step further to provide easy-to-use, round-the-clock access to spatial data and decision support over the Web. Traditional DSS and Web-based spatial DSS can be further improved by integrating expert knowledge and utilizing intelligent software components (such as expert systems and intelligent agents) to emulate the human intelligence and decision making. These kinds of decision support systems are classified as intelligent decision support systems. The objective of this chapter is to discuss the development of an intelligent web-based spatial decision support system and demonstrate it with a case study for planning snow removal operations.

ObjectiveThe study investigates users’ tendency to access decision support (DS) systems as a function of the correlation between the DS information and the information users already have, the ongoing interaction with such systems, and the effect of correlated information on subjective trust.BackgroundPrevious research has shown inconclusive findings regarding whether people prefer information that correlates with information they already have. Some studies conclude that individuals recognize the value of noncorrelated information, given its unique content, while others suggest that users favor correlated information as it aligns with existing evidence. The impact of the level of correlation on performance, subjective trust, and the decision to use DS remains unclear.MethodIn an experiment (N = 481), participants made classification decisions based on available information. They could also purchase additional DS with different degrees of correlation with the available information.ResultsParticipants tended to purchase information more often when the DS was not correlated with the available information. Correlated information reduced performance, and the effect of correlation on subjective trust and performance depended on DS sensitivity.ConclusionAdditional information may not improve performance when it is correlated with available information (i.e., it is redundant). Hence, the benefits of additional information and DS depend on the information the system and the operator use.ApplicationIt is essential to analyze the correlations between information sources and design the available information to allow optimal task performance and possibly minimize redundancy (e.g., by locating sensors in different positions to capture independent data).

Background: Heart diseases continue to be among the primary causes of morbidity and mortality worldwide, highlighting the need for novel diagnostic methods to accurately, early, and economically predict and diagnose heart diseases. Current physical explores upon expert knowledge, possibly missing intricate and non-linear associations between heart disease risk factors. This is where machine learning (ML)methods emerge as promising alternatives for prediction and diagnosis of heart diseases by emphasizing data availability. Methods: This paper attempts to find out how six popular machine learning algorithms namely, Logistic Regression (LR), Support Vector Machine (SVM), Random Forest (RF), Gradient Boosting (GB), K-Nearest Neighbors (KNN), and eXtreme Gradient Boosting (XGB) perform on a publicly available heart disease dataset collected from Kaggle. For better performance and to make them comparable to each other, basic preprocessing techniques have also been implemented. The performance of all these algorithms will be measured by their respective definitions of accuracy, precision, recall, and F1 score. Result: With an F1-score of 98.52%, Random Forest and XGBoost surpassed other models, yielding a superior accuracy in predictions, measuring 98.53%. The models recorded a perfect level of precision, 100%, coupled with high values of recall, exceeding 97%. Gradient Boosting recorded impressive results, receiving an accuracy score of 93.17%. Conclusion: The research findings revealed that the Random Forest and XGBoost models are very effective in the prediction and diagnosis of heart disease. These two models have the capacity to support a clinical decision. Novelty: That is to say, the novelty of this research is that it evaluates multiple machine learning models with comparison to each other. The role of machine learning in enhancing cardiovascular disease diagnosis as well as decision support systems in healthcare and educating and aware the people about heart care and health.

This paper examines how an object called a decision support system (DSS) is a system from the point of view of GST (general systems theory) and how GST is related with the development of DSS in a positive way. Since a DSS is a complicated system, it cannot be described by one model. We will proceed the consideration in the following different levels: functional level, structural level, process level and physical level. We will show that on each level GST provides a useful framework to deal with a DSS and the general formulation by GST directs what is to be done in the development of DSS. On the progress of the development we have found GST can help us with conceptualizing a DSS which is most important in finding a correct research direction. The presented models are the basic framework of our current system, a product of GST.

Clinical decision support software (CDSS) has been increasingly implemented to assist improved prescribing practice. Reviews and studies report generally positive results regarding prescribing changes and, to a lesser extent, patient outcomes. Little information is available, however, concerning the use of CDSS in community pharmacy practice. Given the apparent paucity of publications examining this topic, we conducted a review to determine whether CDSS in community pharmacy practice can improve medication use and patient outcomes. A literature search of articles on CDSS relevant to community pharmacy and published between 1 January 2005 and 21 October 2013 was undertaken. Articles were included if the healthcare setting was community pharmacy and the article indicated that pharmacy use of CDSS was part of the study intervention. Eight studies were found which assessed counselling, selected drug interactions, inappropriate prescribing and under-prescribing. One study was halted due to insufficient data collection. Six studies showed statistically significant improvements in the measured outcomes: increased patient counselling, 31% reduced frequency of drug-drug interactions (DDIs), reduced frequency of inappropriate medications in the elderly (2·2-1·8% patients) and in pregnant women (5·5-2·9% patients), and increased pharmacists' interventions for under-prescribed low-dose aspirin (1·74 vs. 0·91 per 100 patients with type 2 diabetes) and over-prescribed high-dose proton-pump inhibitors (PPIs) (1·67 vs. 0·17 interventions per 100 high-dose PPI prescriptions). Most studies showed improved prescribing practice, via direct communication between pharmacists and doctors or indirectly via patient education. Factors limiting the impact of improved prescribing included alert fatigue and clinical inertia. No study investigated patient outcomes and little investigation had been undertaken on how CDSS could be best implemented. Few studies have been undertaken in community pharmacy practice, and based on the positive findings reported, further research should be directed in this area, including investigation of patient outcomes.

The growth in the number of cloud computing users has led to the availability of a variety of cloud based services provided by different vendors. This has made the task of selecting suitable set of services quite difficult. There has been a lot of research towards the development of suitable decision support system (DSS) to assist users in making an optimal selection of cloud services. However, existing decision support systems cannot address two crucial issues: firstly, the involvement of both business and technical perspectives in decision making simultaneously and, secondly, the multiple-clouds services based selection using single DSS. In this paper, we tackle these issues in the light of solving the problem of cloud service discovery. In particular, we present the following novel contributions: Firstly, we present critical analysis of the state-of-the-art in decision support systems. Based on our analysis, we identify critical shortcomings in the existent tools and develop the set of requirements which should be met by a potential DSS. Secondly, we present a new holistic framework for the development of DSS which allows a pragmatic description of user requirements. Additionally, the data gathering and analysis is studied as an integral part of the proposedDSS and therefore, we present concrete algorithms to assess the data for an optimal service discovery. Thirdly, we assess our framework for applicability to cloud service selection using an industrial case study. We also demonstrate the implementation and performance of our proposed framework using a prototype which serves as a proof of concept. Overall, this paper provides novel and holistic framework for development of a multiple cloud service discovery based decision support system.

The increasing frequency of disasters highlights the critical need for efficient and adaptable humanitarian logistics systems.To enhance decision-making in these complex environments, it is essential to incorporate advanced mathematical models through the development of Decision Support Systems (DSS).This study presents the development of a DSS prototype aimed at optimizing the location of temporary Points of Distribution (PODs) following a large-scale earthquake scenario in Alto Selva Alegre, Arequipa, Peru.The system distinguishes between two types of PODs: fixed PODs, which remain operational throughout all periods and serve as stable infrastructure during both the response and recovery phases, and temporary PODs, which are activated only during the initial response to maximize proximity to the affected population.The DSS is composed of four core components-a database, an optimization model, a user interface, and a communication layer-and is structured into five interactive tabs. The Data tab presents input information; Optimization tab enables model execution and parameter adjustments based on decision-maker knowledge, resource availability, or operational priorities; Solution visualizes results by period; andComparative allows users to analyze, and contrast saved configurations.Results show that small changes in input parameters-such as the number of modules or active PODs-can lead to significant improvements in coverage and travel distance, with only modest increases in cost.The DSS also adapts well to operational constraints, providing feasible solutions even when some sites are unavailable.By combining mathematical modelling with interactive usability, it supports quick scenario comparison and informed decision-making, offering a practical tool for humanitarian logistics planning.