In search of deterministic methods for initializing K-means and Gaussian mixture clustering

نویسندگان

  • Ting Su
  • Jennifer G. Dy
چکیده

The performance of K-means and Gaussian mixture model (GMM) clustering depends on the initial guess of partitions. Typically, clus∗corresponding author

برای دانلود رایگان متن کامل این مقاله و بیش از 32 میلیون مقاله دیگر ابتدا ثبت نام کنید

ثبت نام

اگر عضو سایت هستید لطفا وارد حساب کاربری خود شوید

منابع مشابه

Convolutional density estimation in hidden Markov models for speech recognition

In continuous density Hidden Markov Models (HMMs) for speech recognition, the probability density function (pdf) for each state is usually expressed as a mixture of Gaussians. In this paper, we present a model in which the pdf is expressed as the convolution of two densities. We focus on the special case where one of the convolved densities is a M -Gaussian mixture, and the other is a mixture o...

متن کامل

The EM Algorithm used for Gaussian Mixture Modelling and its Initialization

Initialization Patricia McKenzie and Michael Alder Center for Intelligent Information Processing Systems The University of Western Australia Nedlands W.A. 6009, AUSTRALIA Abstract We look at the EM algorithm used for Gaussian Mixture Modelling and problems with its initialization. We then consider a method of initializing the algorithm to cluster centers called the Dog Rabbit Strategy and compa...

متن کامل

A Hybrid Data Clustering Algorithm Using Modified Krill Herd Algorithm and K-MEANS

Data clustering is the process of partitioning a set of data objects into meaning clusters or groups. Due to the vast usage of clustering algorithms in many fields, a lot of research is still going on to find the best and efficient clustering algorithm. K-means is simple and easy to implement, but it suffers from initialization of cluster center and hence trapped in local optimum. In this paper...

متن کامل

GROUND MOTION CLUSTERING BY A HYBRID K-MEANS AND COLLIDING BODIES OPTIMIZATION

Stochastic nature of earthquake has raised a challenge for engineers to choose which record for their analyses. Clustering is offered as a solution for such a data mining problem to automatically distinguish between ground motion records based on similarities in the corresponding seismic attributes. The present work formulates an optimization problem to seek for the best clustering measures. In...

متن کامل

An Improved K-Means with Artificial Bee Colony Algorithm for Clustering Crimes

Crime detection is one of the major issues in the field of criminology. In fact, criminology includes knowing the details of a crime and its intangible relations with the offender. In spite of the enormous amount of data on offenses and offenders, and the complex and intangible semantic relationships between this information, criminology has become one of the most important areas in the field o...

متن کامل

ذخیره در منابع من


  با ذخیره ی این منبع در منابع من، دسترسی به آن را برای استفاده های بعدی آسان تر کنید

عنوان ژورنال:
  • Intell. Data Anal.

دوره 11  شماره 

صفحات  -

تاریخ انتشار 2007