Volume 60, Issue 4 - Data-Driven Decision Support / Guest Editor: Stefan Thalmann, Host Editor: Ronald Maier

Due to an ever-increasing amount of data generated in healthcare each day, healthcare professionals are more and more challenged with information. Predictive models based on machine learning algorithms can help to quickly identify patterns in clinical data. Requirements for data driven decision support systems for health and care ( DS4H ) are similar in many ways to applications in other domains. However, there are also various challenges which are specific to health and care settings. The present paper describes a) healthcare specific requirements for DS4H and b) how they were addressed in our Predictive Analytics Toolset for Health and care ( PATH ). PATH supports the following process: objective definition, data cleaning and pre-processing, feature engineering, evaluation, result visualization, interpretation and validation and deployment. The current state of the toolset already allows the user to switch between the various involved levels, i. e. raw data (ECG), pre-processed data (averaged heartbeat), extracted features (QT time), built models (to classify the ECG into a certain rhythm abnormality class) and outcome evaluation (e. g. a false positive case) and to assess the relevance of a given feature in the currently evaluated model as a whole and for the individual decision. This allows us to gain insights as a basis for improvements in the various steps from raw data to decisions.

Readmission prediction in hospitals is a highly complex task involving multiple risk factors that can vary among different disease groups. We address this issue by implementing multiple cross-validated classification models within an intelligent CDSS to enhance patient discharge management. Depending on the diagnosis, the system selects and applies the appropriate model and visualises the prediction results. In addition, the cost and reimbursement development for each episode are determined. The architecture of the CDSS and the integration of the prediction models are presented in this paper.

As the IT landscape of organizations increasingly needs to comply with various laws and regulations, organizations manage a plethora of security-related data and have to verify the adequacy and effectiveness of their security controls through internal and external audits. Existing Governance, Risk and Compliance (GRC) approaches provide little support for auditors or are tailored to the needs of auditors and do not fully support required management activities of the auditee. To address this gap and move towards a holistic solution, a data-driven approach is proposed. Following the design science research paradigm, a data-driven approach for audit data management and analytics that addresses organizational needs as well as requirements for audit data analytics was developed. We contribute workflow support and associated data models to support auditing and security decision making processes. The evaluation shows the viability of the proposed IT artifact and its potential to reduce costs and complexity of security management processes and IT security audits. By developing a model and associated decision support workflows for the entire IT security audit lifecycle, we present a solution for both the auditee and the auditor. This is useful to developers of GRC tools, vendors, auditors and organizational decision makers.

A challenge for importers in the automobile industry is adjusting to rapidly changing market demands. In this work, we describe a practical study of car import planning based on the monthly car registrations in Austria. We model the task as a data driven forecasting problem and we implement four different prediction approaches. One utilizes a seasonal ARIMA model, while the other is based on LSTM-RNN and both compared to a linear and seasonal baselines. In our experiments, we evaluate the 33 different brands by predicting the number of registrations for the next month and for the year to come.

We describe a case study on decision making in command centers of security forces at major open air music festivals. Our goal was to assess current modus operandi and to identify design implications. We have carried out two expert interviews, two field observations and four group discussions with experts based on a fully functional prototype as IT artifact to concretize discussions. Key insights are that in this case localization of relevant resources is the most important aspect of situation awareness, and that state of current knowledge needs to be clearly shared within the command center.

We present our decision support middleware PROViDE that facilitates decentralized decision making in multi-robot teams operating in highly dynamic environments with potentially unreliable communication channels and noisy sensors. Achieving an adaptive team behavior in such an environment is a challenge because the specific conditions require a fully decentralized decision process. The design of PROViDE borrows inspiration from human decision making processes. PROViDE supports replication of proposals, conflict resolution, and final team-decision making. For each of these steps a choice of methods is offered to the developer to provide flexibility for different application requirements and characteristics of execution environments. PROViDE is integrated into a comprehensive modeling framework for multi-robot systems. The main contributions of this paper are twofold: For the development of adaptive multi-robot teams we discuss requirements for a middleware that supports decentralized decision making in dynamic and adverse environments, and we demonstrate the effective and coherent integration of a set of domain-dependent decision support protocols into a middleware framework.