Browsing by Author "Aydin,M.N."

Now showing 1 - 4 of 4

Citation Count: 0
Analysis and Implications of the Giant Component for an Online Interactive Platform
(International Business Information Management Association, IBIMA, 2016) Aydin,M.N.; Perdahci,N.Z.
This research is concerned with practical and research challenges related to understanding the nature of online interactive platforms. So-called network science is adopted to investigate the very nature of these systems as complex systems. In this regard, we examine an online interactive health network and show that the interactive platform examined exhibits essential structural properties that characterize most real complex networks. We basically look into the largest connected component, so-called a giant component (GC), to better understand how the representative network has established. In particular, we apply dynamic network analysis to investigate how the GC has evolved over time. We identify a particular pattern towards emerging a GC. Implications of the patterns have been elaborated from a management perspective. We recommend that the basic stages of the emergence of the GC might be of interest to platform managers while evaluating performance of online platforms.
Citation Count: 0
A European Perspective on Innovation Management a Semi-Structured Model for the Corporate Innovation System (cis)
(Springer International Publishing, 2023) Pasin,M.; Aydin,M.N.; Ovaci,C.
The new paradigm of digitalization represents disruptive changes for organizations around the world. Companies are facing with highly intense competition. In order to survive and achieve sustainable competition advantage, strategic innovation management becomes essential. In this regard, one of the most significant issues is to design and apply a model that includes a clear roadmap to implement innovation principles and activities to ensure innovation capability and performance of businesses. The first part of the chapter presents state-of-the-art literature on existing innovation management terminologies and models. The other parts provide a semi-structured corporate innovation system (CIS) model and its dimensions. The proposed semi-structured CIS model is articulated in terms of the model dimensions and their instantiations along the rich associated experiences gained via best practices of the successful nationwide innovation program. The proposed CIS is a holistic model that creates value by establishing strategic, cultural, and organizational infrastructure for innovation management. The CIS model provides a roadmap from initial evaluations of innovation performance and strategy formulation to implementation. Besides, the model enables us to customize the roadmap based on six dimensions and 20 key target indicators according to company needs and structure. It is a unique model as it aims to establish a system based on the requirements and readiness of organizations. © The Editor(s) (if applicable) and The Author(s), under exclusive license to Springer Nature Switzerland AG 2024.
Citation Count: 24
A hybrid deep learning framework for unsupervised anomaly detection in multivariate spatio-temporal data
(MDPI AG, 2020) Karadayi,Y.; Aydin,M.N.; Ög˘renci,A.S.
Multivariate time-series data with a contextual spatial attribute have extensive use for finding anomalous patterns in a wide variety of application domains such as earth science, hurricane tracking, fraud, and disease outbreak detection. In most settings, spatial context is often expressed in terms of ZIP code or region coordinates such as latitude and longitude. However, traditional anomaly detection techniques cannot handle more than one contextual attribute in a unified way. In this paper, a new hybrid approach based on deep learning is proposed to solve the anomaly detection problem in multivariate spatio-temporal dataset. It works under the assumption that no prior knowledge about the dataset and anomalies are available. The architecture of the proposed hybrid framework is based on an autoencoder scheme, and it is more efficient in extracting features from the spatio-temporal multivariate datasets compared to the traditional spatio-temporal anomaly detection techniques. We conducted extensive experiments using buoy data of 2005 from National Data Buoy Center and Hurricane Katrina as ground truth. Experiments demonstrate that the proposed model achieves more than 10% improvement in accuracy over the methods used in the comparison where our model jointly processes the spatial and temporal dimensions of the contextual data to extract features for anomaly detection. © 2020 by the authors.
Citation Count: 1
School-wide friendship metadata correlations
(Elsevier Ltd, 2019) Aydin,M.N.; Perdahci,Z.N.
Managers and education practitioners desire to know an extent to which sustainable school-wide friendship exists. Drawing on theory of network, this research focuses on bestfriendships that may contribute to positive school experience or school belonging in the context of school-wide interactions. We emphasize that school-wide unity is essential to refer to shared perceived friendship experience at the school level. The basic trust of this study is that managers should consider interconnectedness as a complex system of entangled interactions among students. We investigate best friendship network on the meso-to-macro scale. Particular attention is paid to the network phenomena of the largest component and network correlations for examining school-wide unity. The results show that abundance of asymmetric friendships leads to unity around school wide interactions. As suggested by network theory, popular students’ tendency to avoid forming closed clusters assures sustainability in school-wide friendships, and having same gender type or being classmates correlate highly with the choice of best friends, in contrast to achievement scores. Metadata correlations reveal same-gender and same-class clubs. Incorporating meso level findings into macro level indicates that some metadata (e.g. gender) may be considered as salient characteristics of the communities while other metadata (e.g. achievement scores) may be irrelevant. © 2018 Elsevier Ltd