Refine
Year of publication
Document Type
- Conference Proceeding (492)
- Article (218)
- Part of a Book (48)
- Doctoral Thesis (31)
- Other Publications (28)
- Master's Thesis (14)
- Report (13)
- Working Paper (12)
- Book (9)
- Bachelor Thesis (8)
Language
- English (883) (remove)
Keywords
- (Strict) sign-regularity (1)
- 1D-CNN (1)
- 2 D environment Laser data (1)
- 360-degree coverage (1)
- 3D Extended Object Tracking (1)
- 3D Extended Object Tracking (EOT) (2)
- 3D shape tracking (1)
- 3D ship detection (1)
- 3D urban planning (1)
- AAL (3)
Institute
- Fakultät Architektur und Gestaltung (6)
- Fakultät Bauingenieurwesen (26)
- Fakultät Elektrotechnik und Informationstechnik (16)
- Fakultät Informatik (63)
- Fakultät Maschinenbau (12)
- Fakultät Wirtschafts-, Kultur- und Rechtswissenschaften (43)
- Institut für Angewandte Forschung - IAF (78)
- Institut für Optische Systeme - IOS (36)
- Institut für Strategische Innovation und Technologiemanagement - IST (38)
- Institut für Systemdynamik - ISD (98)
“Crowd contamination”?
(2023)
Misconduct allegations have been found to not only affect the alleged firm but also other, unalleged firms in form of reputational and financial spillover effects. It has remained unexplored, however, how the number of prior allegations against other firms matters for an individual firm currently facing an allegation. Building on behavioral decision theory, we argue that the relationship between allegation prevalence among other firms and investor reaction to a focal allegation is inverted U-shaped. The inverted U-shaped effect is theorized to emerge from the combination of two effects: In the absence of prior allegations against other firms, investors fail to anticipate the focal allegation, and hence react particularly negatively (“anticipation effect”). In the case of many prior allegations against other firms, investors also react particularly negatively because investors perceive the focal allegation as more warranted (“evaluation effect”). The multi-industry, empirical analysis of 8,802 misconduct allegations against US firms between 2007 and 2017 provides support for our predicted, inverted U-shaped effect. Our study complements recent misconduct research on spillover effects by highlighting that not only a current allegation against an individual firm can “contaminate” other, unalleged firms but that also prior allegations against other firms can “contaminate” investor reaction to a focal allegation against an individual firm.
Domain-specific modeling is increasingly adopted by the software development industry. While textual domain-specific languages (DSLs) already have a wide impact, graphical DSLs still need to live up to their full potential. Textual DSLs are usually generated from a grammar or other short textual notations; their development is often cost-efficient. In this paper, we describe an approach to similarly create graphical DSLs from textual notations. The paper describes an approach to generate a graphical node and edge online editor, using a set of carefully designed textual DSLs to fully describe graphical DSLs. Combined with an adequate metamodel, these textual definitions represent the input for a generator that produces a graphical Editor for the web with features such as collaboration, online storage and being always available. The entire project is made available as open source under the name Zeta. This paper focuses on the overall approach and the description of the textual DSLs that can be used to develop graphical modeling languages and editors.
Domain-specific modeling is increasingly adopted in the software development industry. While textual domain-specific languages (DSLs) already have a wide impact, graphical DSLs still need to live up to their full potential. In this paper, we describe an approach to automatically generate a graphical DSL from a set of textual languages. With our approach, node and edge type graphical DSLs can be described using textual models. A set of carefully designed textual DSLs is the input for our generators. The result of the generation is a graphical editor for the intended domain. The development time for a graphical editor is reduced significantly. The whole project is available as open source under the name "Zeta". This publication focuses on the explanation of the textual DSLs for defining a graphical node and edge editor.
This work studies a wind noise reduction approach for communication applications in a car environment. An endfire array consisting of two microphones is considered as a substitute for an ordinary cardioid microphone capsule of the same size. Using the decomposition of the multichannel Wiener filter (MWF), a suitable beamformer and a single-channel post filter are derived. Due to the known array geometry and the location of the speech source, assumptions about the signal properties can be made to simplify the MWF beamformer and to estimate the speech and noise power spectral densities required for the post filter. Even for closely spaced microphones, the different signal properties at the microphones can be exploited to achieve a significant reduction of wind noise. The proposed beamformer approach results in an improved speech signal regarding the signal-to-noise-ratio and keeps the linear speech distortion low. The derived post filter shows equal performance compared to known approaches but reduces the effort for noise estimation.
Nowadays established companies use Corporate Entrepreneurship (CE) as a means to create discontinuous innovations. Many companies thereby even implement multiple CE units that typically involve several entrepreneurial activities. This explorative study aimed to identify the reasons why established companies implement multiple CE units concurrently. In conducting a comparative case study with eight companies from different industries, valuable insights for science and practice were gained. We provide an overview of different 11 reasons for implementing multiple CE units. This shows that the combination of CE units used by companies differs depending on the reason. It further allowed to derive general approaches of established companies to the implementation of CE units. Last, we identify the concept of co-specialization to be a central driver explaining the creation of the need to set up multiple units. We conclude by indicating implications and subjects for future research.
The cornerstone of cognitive systems is environment awareness which enables agile and adaptive use of channel resources. Whitespace prediction based on learning the statistics of the wireless traffic has proven to be a powerful tool to achieve such awareness. In this paper, we propose a novel Hidden Markov Model (HMM) based spectrum learning and prediction approach which accurately estimates the exact length of the whitespace in WiFi channels within the shared industrial scientific medical ISM) bands. We show that extending the number of hidden states and formulating the prediction problem as a maximum likelihood (ML) classification leads to a substantial increase in the prediction horizon compared to classical approaches that predict the immediate (short-term) future. We verify the proposed algorithm through simulations which utilize a model for WiFi traffic based on extensive measurement campaigns.
Times of high dynamic and growing new knowledge demand for entrepreneurial education and university engagement. Higher education institutions (HEIs) have established intensive knowledge and resources about entrepreneurial education and relating activities and formats over the last years. As smaller companies (SMEs) are increasingly experimenting with entrepreneurship, they seem to struggle with setting up entrepreneurial activities within their established corporate strategy and innovation structures. It is beneficial for them to collaborate with higher education institutions to minimize the risk and uncertainty associated with implementing entrepreneurship education (EE) and catch up with larger corporates. Further, research lacks a systematic characterization of EE activities in those companies and classification of collaboration formats. Therefore, this study uses qualitative research methods to analyze data from interviews conducted with two German SMEs. Our study contributes to a better understanding of EE in SME and respective HEI collaborations by (1) characterizing EE in SME and SME-HEI collaboration based on attributes and collaboration types defined by their locus of collaboration and intensity of knowledge inflow and (2) identifying differences among EE in SME and HEI. We provide implications to practice—corporate and university EE initiatives—for a more effective design and implementation of EE in SMEs and the SME-HEI collaborations themselves.
Volterra and Wiener series
(2011)
Volterra and Wiener series are two classes of polynomial representations of nonlinear systems. They are perhaps the best understood and most widely used nonlinear system representations in signal processing and system identification. A Volterra or Wiener representation can be thought of as a natural extension of the classical linear system representation. In addition to the convolution of the input signal with the system's impulse response, the system representation includes a series of nonlinear terms that contain products of increasing order of the input signal with itself. It can be shown that these polynomial extension terms allow for representing a large class of nonlinear systems which basically encompasses all systems with scalar outputs that are time-invariant and have noninfinite memory.
Visualization-Assisted Development of Deep Learning Models in Offline Handwriting Recognition
(2018)
Deep learning is a field of machine learning that has been the focus of active research and successful applications in recent years. Offline handwriting recognition is one of the research fields and applications were deep neural networks have shown high accuracy. Deep learning models and their training pipeline show a large amount of hyper-parameters in their data selection, transformation, network topology and training process that are sometimes interdependent. This increases the overall difficulty and time necessary for building and training a model for a specific data set and task at hand. This work proposes a novel visualization-assisted workflow that guides the model developer through the hyper-parameter search in order to identify relevant parameters and modify them in a meaningful way. This decreases the overall time necessary for building and training a model. The contributions of this work are a workflow for hyper-parameter search in offline handwriting recognition and a heat map based visualization technique for deep neural networks in multi-line offline handwriting recognition. This work applies to offline handwriting recognition, but the general workflow can possibly be adapted to other tasks as well.