scholarly journals Cluster analysis of knowledge sources in standardized electrical engineering subfields

2016 ◽  
Vol 13 (3) ◽  
pp. 405-422
Author(s):  
Marija Blagojevic ◽  
Zivadin Micic ◽  
Momcilo Vujicic
Keyword(s):  
Neural Networks ◽  
Cluster Analysis ◽  
Electrical Engineering ◽  
Financial Resources ◽  
Knowledge Sources ◽  
Knowledge Source ◽  
Innovation Intensity ◽  
Future Work ◽  
Analysis Of Knowledge ◽  
Mathematical Relations

The paper presents a cluster analysis of innovation of knowledge sources based on the standards in the field of Electrical Engineering. Both local (SRPS) and global (ISO) knowledge sources have been analysed with the aim of innovating a Knowledge Base (KB). The results presented indicate a means/possibility of grouping the subfields within a cluster. They also point to a trend or intensity of knowledge source innovation for the purpose of innovating the KB that accompanies innovations. The study provides the possibility of predicting necessary financial resources in the forthcoming period by means of original mathematical relations. Furthermore, the cluster analysis facilitates the comparison of the innovation intensity in this and other (sub)fields. Future work relates to the monitoring of the knowledge source innovation by means of KB engineering and improvement of the methodology of prediction using neural networks.

Exploring the Unknown Nature of Data

2010 ◽  
pp. 1-27
Author(s):  
Rui Xu ◽  
Donald C. Wunsch II
Keyword(s):  
Neural Networks ◽  
Cluster Analysis ◽  
Data Structures ◽  
Clustering Algorithms ◽  
Ground Truth ◽  
Human Beings ◽  
Data Set ◽  
Learning Capabilities ◽  
Good Learning ◽  
Hidden Data

To classify objects based on their features and characteristics is one of the most important and primitive activities of human beings. The task becomes even more challenging when there is no ground truth available. Cluster analysis allows new opportunities in exploring the unknown nature of data through its aim to separate a finite data set, with little or no prior information, into a finite and discrete set of “natural,” hidden data structures. Here, the authors introduce and discuss clustering algorithms that are related to machine learning and computational intelligence, particularly those based on neural networks. Neural networks are well known for their good learning capabilities, adaptation, ease of implementation, parallelization, speed, and flexibility, and they have demonstrated many successful applications in cluster analysis. The applications of cluster analysis in real world problems are also illustrated. Portions of the chapter are taken from Xu and Wunsch (2008).

Application of the Cluster Analysis in Computational Paleography

2016 ◽  
pp. 525-543
Author(s):  
Loránd Lehel Tóth ◽  
Raymond Eliza Ivan Pardede ◽  
György András Jeney ◽  
Ferenc Kovács ◽  
Gábor Hosszú
Keyword(s):  
Cluster Analysis ◽  
Unknown Origin ◽  
Canonical Decomposition ◽  
Minkowski Metric ◽  
Clustering Technique ◽  
Future Work

This chapter presents a method to determine the actual version of a script used in constructing of a script relic from unknown origin. The glyphs belong to graphemes as models are realized in the relics as symbols. Some group of glyphs may transform their shape (shapeshifting) through time which produces various versions of scripts that use different glyphs to express the same grapheme. These glyph variants can be identified from extant relics, mainly from historical abecedaries that are used as references. Our algorithm can determine whether or not an abecedary is related to the symbols of a relic from unknown origin by means of the canonical decomposition of the glyphs and symbols. From there an aggregated value called fingerprint is created and it is unique for each relic. The fingerprints then are evaluated by clustering technique using various metrics. As the result of performing comparative evaluations the Minkowski metric provides the most interpretable clustering structure. The results of the evaluations, conclusions, and future work are also presented.

Web-Based Cooperative Information Systems Modeling

2011 ◽  
pp. 193-206
Author(s):  
Youcef Baghdadi
Keyword(s):  
Information System ◽  
Information Systems ◽  
External Information ◽  
Knowledge Sources ◽  
Knowledge Source ◽  
Web Based ◽  
Cooperative Information Systems ◽  
Business Objects ◽  
Interaction Component ◽  
The Web

This chapter presents a modeling for the Web-Based Cooperative Information Systems (WBCISs). This modeling considers the WBCISs as support of the unavoidable interactions among multiple existing heterogeneous subsystems of the information system and external information sources that share business objects and processes. The WBCIS is considered as an artifact that firstly and mainly allow information exchange, coordination and cooperation among these sources; and secondly data restructing and processes reuse or reengineering. The main concepts are Knowledge Sources and Interaction Component. A knowledge source represents a subsystem of the information system (Personal IS, Workgroup IS or Enterprise IS); or any external information source. The modeling considers the knowledge source as a UML package that presents an interface definition (business objects schema and processes). An interactions component is a kind of Web-based broker of business objects and processes. It is a support for communication services and user-oriented semantic services of the knowledge sources. It is based on the Web so that it deals with semi-structured data and accesses any knowledge source (willing to interact) having its URL. It uses a metadata that describes the knowledge sources as UML package. The modeling specializes interactions components according to the interaction situations of the knowledge source namely interactions for coordination that deal with the consistency of the shared business objects, interactions for cooperation related to the coupled processes’ activities or interactions for transmission that deal with informal and unstructured information exchanges. A Coordination Component allows knowledge source location, access, integration, global view and restructing of the business objects. A Cooperation Component allows process’ activities invocation, reuse or reengineering activities. This methodologic specialization allows easier implementation and reuse of the interaction components. An interaction component is modeled as a UML package.

Classification of Fourier Transform Infrared Microscopic Imaging Data of Human Breast Cells by Cluster Analysis and Artificial Neural Networks

2003 ◽  
Vol 57 (1) ◽  
pp. 14-22 ◽  
Author(s):  
Lin Zhang ◽  
Gary W. Small ◽  
Abigail S. Haka ◽  
Linda H. Kidder ◽  
E. Neil Lewis
Keyword(s):  
Neural Networks ◽  
Cluster Analysis ◽  
Fourier Transform ◽  
Artificial Neural Networks ◽  
Fourier Transform Infrared ◽  
Carcinoma Cells ◽  
Human Breast ◽  
Microscopic Imaging ◽  
Artificial Neural

Cluster analysis and artificial neural networks (ANNs) are applied to the automated assessment of disease state in Fourier transform infrared microscopic imaging measurements of normal and carcinomatous immortalized human breast cell lines. K-means clustering is used to implement an automated algorithm for the assignment of pixels in the image to cell and non-cell categories. Cell pixels are subsequently classified into carcinoma and normal categories through the use of a feed-forward ANN computed with the Broyden–Fletcher–Goldfarb–Shanno training algorithm. Inputs to the ANN consist of principal component scores computed from Fourier filtered absorbance data. A grid search optimization procedure is used to identify the optimal network architecture and filter frequency response. Data from three images corresponding to normal cells, carcinoma cells, and a mixture of normal and carcinoma cells are used to build and test the classification methodology. A successful classifier is developed through this work, although differences in the spectral backgrounds between the three images are observed to complicate the classification problem. The robustness of the final classifier is improved through the use of a rejection threshold procedure to prevent classification of outlying pixels.

On-line determination of ore type using cluster analysis and neural networks

1995 ◽  
Vol 8 (6) ◽  
pp. 637-648 ◽  
Author(s):  
S Laine ◽  
H Lappalainen ◽  
S.-L Jämsä-Jounela
Keyword(s):  
Neural Networks ◽  
Cluster Analysis ◽  
On Line

scholarly journals Meta-Heuristic Optimization Methods for Quaternion-Valued Neural Networks

Mathematics ◽  
2021 ◽  
Vol 9 (9) ◽  
pp. 938
Author(s):  
Jeremiah Bill ◽  
Lance Champagne ◽  
Bruce Cox ◽  
Trevor Bihl
Keyword(s):  
Neural Networks ◽  
Optimization Methods ◽  
Network Size ◽  
High Dimensional ◽  
Proof Of Concept ◽  
Activation Functions ◽  
Hypercomplex Numbers ◽  
Neural Network Structure ◽  
Future Work ◽  
High Dimensional Datasets

In recent years, real-valued neural networks have demonstrated promising, and often striking, results across a broad range of domains. This has driven a surge of applications utilizing high-dimensional datasets. While many techniques exist to alleviate issues of high-dimensionality, they all induce a cost in terms of network size or computational runtime. This work examines the use of quaternions, a form of hypercomplex numbers, in neural networks. The constructed networks demonstrate the ability of quaternions to encode high-dimensional data in an efficient neural network structure, showing that hypercomplex neural networks reduce the number of total trainable parameters compared to their real-valued equivalents. Finally, this work introduces a novel training algorithm using a meta-heuristic approach that bypasses the need for analytic quaternion loss or activation functions. This algorithm allows for a broader range of activation functions over current quaternion networks and presents a proof-of-concept for future work.

Sign in / Sign up

Export Citation Format

Share Document