Using Self-Play within Deep Q Learning to improve real-time Production Scheduling

2021 | Americas Conference on Information Systems | Citations: 0

Authors: Groth, Michael; Freier, Pascal; Schumann, Matthias

Abstract: Current methods of generating production schedules are either time-consuming or ... Expand

Abstract: Current methods of generating production schedules are either time-consuming or do not meet all productivity goals. A solution to this is to facilitate machine learning to generate production schedules. The necessary data to use machine learning is provided by the increasing use of cyber-physical systems in production environments. However, the mere acquisition of mass data through cyber-physical systems does not inherently lead to improvements. The algorithms must learn from historical situations to improve future calculations. Therefore, we implement a machine learning algorithm for real-time production scheduling in this paper. For this, we identify a suitable machine learning algorithm and meta-requirements based on a design science research approach. Finally, we evaluate the created artifact in a simulation study. The results show that the implemented solution can create schedules in real-time, which are of higher quality compared to priority rules. Collapse

Topics: decision support system Pascal

Methods: computational algorithm design artifact design science simulation machine learning

Introduction to Blockchain Cases and Innovations

2021 | HICSS | Citations: 0

Authors: Themistocleous, Marinos; Cunha, Paulo Rupino Da; Christodoulou, Klitos

Abstract: ... Expand

Abstract: Collapse

Topics: blockchain Xing technology adoption business process automation smart contract

Methods: case study literature sample design science

Theories: institutional theory theory of everything

Bayesian neural networks for flight trajectory prediction and safety assessment

2020 | Decision Support Systems | Citations: 4

Authors: Zhang, Xiaoge; Mahadevan, Sankaran

Abstract: Safety, as the most important concern in civil aviation, needs to be maintained ... Expand

Abstract: Safety, as the most important concern in civil aviation, needs to be maintained at an acceptable level at all times in the air transportation system. This paper aims to increase en-route flight safety through the development of deep learning models for trajectory prediction, where model prediction uncertainty is characterized following a Bayesian approach. The proposed methodology consists of four steps. In the first step, a large volume of raw messages in Flight Information Exchange Model (FIXM) format streamed from Federal Aviation Administration are processed with a distributed computing engine Apache Spark to extract trajectory information in an efficient manner. In the second step, two types of deep learning models are trained to predict flight trajectory from different perspectives. Specifically, deep feedforward neural networks (DNN) are trained to make a one-step-ahead prediction on the deviation along latitude and longitude between the actual flight trajectory and target flight trajectory. In parallel, deep Long Short-Term Memory (LSTM) neural networks are trained to make longer-term predictions on the flight trajectory over several subsequent time instants. The DNN model is more accurate but has a single-step prediction horizon, whereas the LSTM model is less accurate but longer prediction horizon. Therefore, in the third step, the two different types of deep learning models are blended together to create a multi-fidelity prediction. After quantifying the discrepancy between the two model predictions in the current time instant, the DNN prediction is used to correct the LSTM prediction of flight trajectory along subsequent time instants accordingly. The multi-fidelity approach is expanded to multiple flights, and is then used to assess safety based on horizontal and vertical separation distance between two flights. Computational results illustrate the promising performance of the blended model in predicting the flight trajectory and assessing en-route flight safety. Collapse

Topics: Apache Spark high-performance computing distributed computing decision making decision support system

Methods: deep learning artificial neural network descriptive statistic computational algorithm machine learning

Customization of IBM Intu’s Voice by Connecting Text-to-Speech Services and a Voice Conversion Network

2018 | HICSS | Citations: 0

Authors: Song, Jongyoon; Kim, Hyunjae; Lee, Jaekoo; Choi, Euishin; Kim, Minseok; Yoon, Sungroh

Abstract: IBM has recently launched Project Intu, which extends the existing web-based cog ... Expand

Abstract: IBM has recently launched Project Intu, which extends the existing web-based cognitive service Watson with the Internet of Things to provide an intelligent personal assistant service. We propose a voice customization service that allows a user to directly customize the voice of Intu. The method for voice customization is based on IBM Watson’s text-to-speech service and voice conversion model. A user can train the voice conversion model by providing a minimum of approximately 100 speech samples in the preferred voice (target voice). The output voice of Intu (source voice) is then converted into the target voice. Furthermore, the user does not need to offer parallel data for the target voice since the transcriptions of the source speech and target speech are the same. We also suggest methods to maximize the efficiency of voice conversion and determine the proper amount of target speech based on several experiments. When we measured the elapsed time for each process, we observed that feature extraction accounts for 59.7% of voice conversion time, which implies that fixing inefficiencies in feature extraction should be prioritized. We used the mel-cepstral distortion between the target speech and reconstructed speech as an index for conversion accuracy and found that, when the number of target speech samples for training is less than 100, the general performance of the model degrades. Collapse

Topics: conversational agent Microsoft Windows user experience Pascal internet of things

Methods: experiment artificial neural network synthesis computational algorithm design artifact

A decision support system for the collaborative selection of strategies in enterprise networks

2016 | Decision Support Systems | Citations: 0

Authors: Andres, Beatriz; Poler, Raul

Abstract: Collaborative networks (CN) consist of autonomous and heterogeneous partners, an ... Expand

Abstract: Collaborative networks (CN) consist of autonomous and heterogeneous partners, and each defines its own objectives and formulates its own strategies, which are selected and activated to achieve these objectives. The heterogeneity that characterises network partners could lead to contradictions appearing among the strategies formulated in each CN enterprise. Consequently, the strategies formulated in one enterprise could negatively influence the achievement of the objectives defined in other enterprises of the same network. These contradictions lead to strategies misalignments, which worsens the network performance. In order to deal with these misalignments, a DSS is proposed to support the process of selecting the strategies among all those formulated, with the aim of achieving higher alignment levels. The proposed DSS considers the impacts that each strategy formulated in each enterprise has on the performance of the objectives defined by each network partner. This allows enterprises to select a set of aligned strategies. The selection of proper strategies to be activated in each enterprise strongly influences the CN's performance level, and higher levels of network adaptability, agility and competitiveness are achieved. The proposed DSS is validated under real conditions in a food industry network. The DSS is evaluated by emulating real collaborative conditions and is compared with the equivalent non-collaborative decision making perspective used for selecting strategies. The results demonstrate that the collaborative approach outperforms the performance level of the non-collaborative one and is more effective for handling the robustness and the long-term operation of the CN. Collapse

Topics: decision support system decision making database system privacy programming language

Methods: case study simulation experiment literature study

A complexity perspective on collaborative decision making in organizations: The ecology of group-performance

2011 | Information & Management | Citations: 0

Authors: Adler, Bernd-Mathias; Baets, Walter; König, Reinhard

Abstract: Networks of communication are essential when managing corporate work and perform ... Expand

Abstract: Networks of communication are essential when managing corporate work and performing information exchange; the systems must allow them to be dynamic and well-structured. They help provide high organizational performance and innovative capacity in today's knowledge intense corporations, and this means that organizations must manage the networks strategically. Despite the fact that practitioners are aware of the huge influence of informal communication on decision making, little is known about the underlying principles of efficient employee network collaboration, which is dynamic in nature, especially for complex environments resulting from steady innovation and high competitive pressure. We addressed this issue from a complexity perspective, using an agent based simulation to visualize the key elements of efficient, information-based, collaborative decision making. Our findings suggested that information and communication technologies (ICT) may not be able to leverage corporate performance of the increasingly complex adaptive organizations. There seem to be elementary natural constraints on the cognitive capacities of people dealing with and managing information. Rather than a better technical approach, a more ecologic one is therefore advocated as the best way to improve decision making. Collapse

Topics: decision making knowledge base object oriented programming language Pascal organizational productivity

Methods: simulation agent based simulation survey cluster analysis sociometry

Theories: social network theory

Reviews

2010 | International Journal of Information Management | Citations: 0

Authors: Fitz-Gerald, Stuart

Abstract: ... Expand

Abstract: Collapse

Topics: Java Prolog Pascal Algol

Methods: design science unified modeling language capability maturity model

Consulting past exceptions to facilitate workflow exception handling

2004 | Decision Support Systems | Citations: 3

Authors: Hwang, San-Yih; Tang, Jian

Abstract: In this paper, we propose an architecture model that deals with both expected an ... Expand

Abstract: In this paper, we propose an architecture model that deals with both expected and unexpected exceptions in the context of workflow management. Expected exceptions and their handling approaches are specified by ECA rules, while cases of unexpected exceptions are characterized by their features and resolution approaches. The handling of unexpected exceptions is then assisted by the system providing information about how recent similar cases were resolved. The ways in which the previous exception cases were handled provides useful information in determining how to handle the current one. Quantifying the similarity of exception cases is described, and three algorithms for efficiently searching for similar exception cases are proposed and evaluated both theoretically and by experimenting with synthetic data sets. Collapse

Topics: decision making Microsoft SQL server database Pascal structured query language knowledge representation

Methods: computational algorithm experiment k-nearest neighbor survey ontological modelling

A Contingency Approach to Software Project Coordination

2001 | Journal of Management Information Systems | Citations: 42

Authors: Andres, Hayward P.; Zmud, Robert W.

Abstract: Before software project managers can enhance productivity and satisfaction of th ... Expand

Abstract: Before software project managers can enhance productivity and satisfaction of the software project team member, the effect of task characteristics, goal orientations, and coordination strategies on design and coding-task outcomes must be understood. A research model, which suggests that task interdependence, goal conflict, and coordination strategies significantly affect productivity and satisfaction associated with software design and coding activities, is presented. Issues such as contingency/design misfit, conflicting contingencies, and the extent of deviation to theoretically prescribed coordination mechanisms applied to contingencies are used to make predictions on productivity and process satisfaction. A 2x2x2 factorial experiment was utilized. Overall, projects characterized by low task interdependence exhibited greater productivity than projects with high task interdependence. Also, in general, organic coordination was more productive than mechanistic coordination. There was also a significant interaction between task interdependence and coordination strategy. Low goal conflict and organic coordination each lead to greater process satisfaction. Productivity results for the goal conflict manipulation was opposite to the hypothesized direction. Unconflicted contingencies addressed with consistent coordination and partially conflicted contingencies, regardless of the coordination used, exhibited significant gains in productivity. In comparison, unconflicted contingencies with inconsistent coordination and conflicted contingencies, regardless of the coordination applied, resulted in lower productivity. This suggests that there are instances where multiple contingencies, which warrant the use of different coordination strategies, can be adequately addressed with a specific coordination strategy. Collapse

Topics: task interdependence productivity systems design system development method systems development

Methods: experiment experimental task laboratory experiment factorial experiment self reported survey

Theories: contingency theory information processing theory

Knowledge Transfer Between Languages and Paradigms

2000 | Americas Conference on Information Systems | Citations: 0

Authors: Armstrong, Deb; Nelson, H James

Abstract: As the field of Information Systems (IS) continues to advance, organizations wil ... Expand

Abstract: As the field of Information Systems (IS) continues to advance, organizations will constantly be facing the challenge of paradigm shifts. This paper extends prior work on IS personnel training by emphasizing the retraining of IS professionals during a paradigm shift. The application of the Osgood (1949) transfer surface to the software development domain is proposed. A model is developed that defines the concepts of positive, negative and no skill transfer within the software development domain. This model demonstrates that the similarity of the stimuli and responses of the software development projects can predict the direction of the skill transfer. Anecdotal evidence of critical points in the model is provided to support applicability to the software development domain. This paper provides a theoretical beginning for work on paradigm shifts in the software development domain and highlights the importance of knowledge transfer in the face of technological change. Collapse

Topics: programming language systems development IT workforce software developer IT development project