Novel Approaches to Unsupervised Clustering Through k-Windows Algorithm

Tasoulis, Dimitris K.; Vrahatis, Michael N.

doi:10.1007/3-540-32394-5_5

Cited by 11 publications

(7 citation statements)

References 48 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…Our clustering algorithm is based on k − windows clustering [6]. The choice of k-windows was made based on the following criteria: (1) The k-windows algorithm is not bound to an initial number of clusters.…”

Section: Clustering Approachmentioning

confidence: 99%

Developing a mobile recommender system

Nakas

Kalogeraki

2012

Proceedings of the 5th International Conference on PErvasive Technologies Related to Assistive Environments

View full text Add to dashboard Cite

The proliferation of powerful, programmable mobile devices along with the availability of wide-area connectivity has created opportunities to sense and share location, motion, acoustic and visual data in a network of mobile devices. In this paper we present a mobile recommender system that exploits the individual data collected by multiple users on their mobile phones. Our aim is to exploit the potential of the computational capabilities of modern mobile devices to provide personalized and better services to the end users. The system has been developed on a MapReduce framework that simplifies the programmability and deployment of the applications on the mobile devices and implements distributed clustering over user rating data. Experimental results over a testbed of Nokia smartphones illustrate the performance and effectiveness of our approach.

show abstract

Section: Clustering Approachmentioning

confidence: 99%

Developing a mobile recommender system

Nakas

Kalogeraki

2012

Proceedings of the 5th International Conference on PErvasive Technologies Related to Assistive Environments

View full text Add to dashboard Cite

show abstract

“…Thus it is essential to reduce the storage overhead by reducing the amount of data retained. Using windows to determine clusters is a plausible technique that allows fast processing, and at the same time is able to produce high quality results [9].…”

Section: Temporal Clustering Though Windowsmentioning

confidence: 99%

“…To this end, as the number of clusters grows more memory is required to keep track of their structure. Moreover, if the clusters have a non-convex shape then more than one window is required to capture their structure [9].…”

Section: The Algorithmmentioning

confidence: 99%

Unsupervised Clustering In Streaming Data

Tasoulis

Adams

Hand

2006

Sixth IEEE International Conference on Data Mining - Workshops (ICDMW'06)

View full text Add to dashboard Cite

Tools for automatically clustering streaming data are becoming increasingly important as data acquisition technology continues to advance. In this paper we present an extension of conventional kernel density clustering to a spatio-temporal setting, and also develop a novel algorithmic scheme for clustering data streams. Experimental results demonstrate both the high efficiency and other benefits of this new approach.

show abstract

“…The recently proposed k-windows clustering algorithm [50] uses a windowing technique to discover the clusters present in an n-dimensional dataset. More specifically, assuming that the dataset lies in n dimensions, the algorithm initializes a number of n-dimensional windows (boxes) over the dataset.…”

Section: The Unsupervised K-windows Clustering Algorithmmentioning

confidence: 99%

“…A fundamental issue in cluster analysis, independent of the particular clustering technique applied, is the determination of the number of clusters present in a dataset. The unsupervised k-windows algorithm is capable to determine the number of clusters through a generalization of the original algorithm [50]. Finally, it must be noted that no objective function evaluations are necessary during the operation of the k-windows clustering algorithm [51].…”

Section: The Unsupervised K-windows Clustering Algorithmmentioning

confidence: 99%

A Review of Major Application Areas of Differential Evolution

Plagianakos

Tasoulis

Vrahatis

Advances in Differential Evolution

Self Cite

View full text Add to dashboard Cite

In this chapter we present an overview of the major applications areas of differential evolution. In particular we pronounce the strengths of DE algorithms in tackling many difficult problems from diverse scientific areas, including single and multiobjective function optimization, neural network training, clustering, and real life DNA microarray classification. To improve the speed and performance of the algorithm we employ distributed computing architectures and demonstrate how parallel, multi-population DE architectures can be utilised in single and multiobjective optimization. Using data mining we present a methodology that allows the simultaneous discovery of multiple local and global minimizers of an objective function. At a next step we present applications of DE in real life problems including the training of integer weight neural networks and the selection of genes of DNA microarrays in order to boost predictive accuracy of classification models. The chapter concludes with a discussion on promising future extensions of the algorithm, and presents novel mutation operators, that are the result of a genetic programming procedure, as very interesting future research direction.

show abstract

Novel Approaches to Unsupervised Clustering Through k-Windows Algorithm

Cited by 11 publications

References 48 publications

Developing a mobile recommender system

Developing a mobile recommender system

Unsupervised Clustering In Streaming Data

A Review of Major Application Areas of Differential Evolution

Contact Info

Product

Resources

About