Results 1  10
of
11
A decomposition algorithm for learning bayesian network structures from data
 in Proceedings of the Twelfth PacificAsia Conference on Knowledge Discovery and Data Mining(PAKDD
"... Abstract. It is a challenging task of learning a large Bayesian network from a small data set. Most conventional structural learning approaches run into the computational as well as the statistical problems. We propose a decomposition algorithm for the structure construction without having to learn ..."
Abstract

Cited by 2 (2 self)
 Add to MetaCart
Abstract. It is a challenging task of learning a large Bayesian network from a small data set. Most conventional structural learning approaches run into the computational as well as the statistical problems. We propose a decomposition algorithm for the structure construction without having to learn the complete network. The new learning algorithm firstly finds local components from the data, and then recover the complete network by joining the learned components. We show the empirical performance of the decomposition algorithm in several benchmark networks.
Sequential Minimal Optimization in AdaptiveBandwidth Convex Clustering
"... Computing not the local, but the global optimum of a cluster assignment is one of the important aspects in clustering. Convex clustering is an approach to acquire the global optimum, assuming some fixed centers and bandwidths of the clusters. The essence of the convex clustering is a convex optimiza ..."
Abstract

Cited by 1 (1 self)
 Add to MetaCart
Computing not the local, but the global optimum of a cluster assignment is one of the important aspects in clustering. Convex clustering is an approach to acquire the global optimum, assuming some fixed centers and bandwidths of the clusters. The essence of the convex clustering is a convex optimization of the mixture weights whose optimum becomes sparse. One of the limitations in the convex clustering was the computational inefficiency of the ExpectationMaximization algorithm, where an extremely large number of iterations is required for the convergence. This paper proposes a more efficient optimization algorithm for convex clustering to significantly reduce the required number of iterations. The key ideas in the proposed algorithm are accurate pruning while choosing a pair of kernels and an elementwise NewtonRaphson method for fast convergence of the nonzero mixture weights. The proposed algorithm is further accelerated when incorporating locally adaptive bandwidths of the clusters, which are primarily introduced to improve the predictive capability. 1
Information Theoretic Clustering using Minimum Spanning Trees
"... Abstract. In this work we propose a new informationtheoretic clustering algorithm that infers cluster memberships by direct optimization of a nonparametric mutual information estimate between data distribution and cluster assignment. Although the optimization objective has a solid theoretical foun ..."
Abstract

Cited by 1 (0 self)
 Add to MetaCart
Abstract. In this work we propose a new informationtheoretic clustering algorithm that infers cluster memberships by direct optimization of a nonparametric mutual information estimate between data distribution and cluster assignment. Although the optimization objective has a solid theoretical foundation it is hard to optimize. We propose an approximate optimization formulation that leads to an efficient algorithm with low runtime complexity. The algorithm has a single free parameter, the number of clusters to find. We demonstrate superior performance on several synthetic and real datasets. 1
Integrating Power line and Wireless for Intelligent and Opportunistic Networking
"... In this paper we try to present a network architecture which leverages the power line for Opportunistic networking. We try to provide the autonomous working by use of 802.11 wireless mesh networks. This architecture is aimed at providing better Quality of Service without any additional infrastructur ..."
Abstract
 Add to MetaCart
In this paper we try to present a network architecture which leverages the power line for Opportunistic networking. We try to provide the autonomous working by use of 802.11 wireless mesh networks. This architecture is aimed at providing better Quality of Service without any additional infrastructure. This integration gives the high bandwidth of wired network to WiFi. The architecture focuses on providing wired backhaul of power line to 802.11s WiFi network. Then the combined network is systematically simplified by graph theory simplification techniques to reduce the network size considerably thereby shortening the route path. It also achieves Quality of Service as each wired backhaul subsequently leads to Internet Gateway. Increasing the backhauls means increase in number of gateways. We devise an intelligent algorithm which can switch between gateways with change in traffic and other overloading conditions thereby bringing the bandwidth within the wireless network at par with its wired counterparts. Categories and Subject Descriptors C.2.1 [Wireless Communication]: Design of wireless LAN, IEEE 802.11s, combining wired and wireless networks, IEEE
Discovering Local Outliers using Dynamic Minimum Spanning Tree with SelfDetection of Best Number of Clusters
"... Detecting outliers in database (as unusual objects) using Clustering and Distancebased approach is a big desire. Minimum spanning tree based clustering algorithm is capable of detecting clusters with irregular boundaries. In this paper we propose a new algorithm to detect outliers based on minimum ..."
Abstract
 Add to MetaCart
Detecting outliers in database (as unusual objects) using Clustering and Distancebased approach is a big desire. Minimum spanning tree based clustering algorithm is capable of detecting clusters with irregular boundaries. In this paper we propose a new algorithm to detect outliers based on minimum spanning tree clustering and distancebased approach. Outlier detection is an extremely important task in a wide variety of application. The algorithm partition the dataset into optimal number of clusters. Small clusters are then determined and considered as outliers. The rest of the outliers (if any) are then detected in the clusters using Distancebased method. The algorithm uses a new cluster validation criterion based on the geometric property of data partition of the dataset in order to find the proper number of clusters. The algorithm works in two phases. The first phase of the algorithm creates optimal number of clusters, where as the second phase of the algorithm detect outliers in the clusters. The key feature of our approach is it combines the best features of Distancebased and Clusteringbased outlier detection to find noisefree/errorfree clusters for a given dataset without using any input parameters.
Minimum Spanning Treebased Clustering Applied to Protein Sequences in Early Cancer Diagnosis
"... Cancer molecular pattern efficient discovery is essential in the molecular diagnostics. The number of amino acid sequence is increasing very rapidly in the protein databases, but the structure of only some amino acid sequences are found in the protein data bank. Thus an important problem in genomics ..."
Abstract
 Add to MetaCart
Cancer molecular pattern efficient discovery is essential in the molecular diagnostics. The number of amino acid sequence is increasing very rapidly in the protein databases, but the structure of only some amino acid sequences are found in the protein data bank. Thus an important problem in genomics is automatically clustering homogeneous protein sequences when only sequence information is available. The characteristics of the protein expression data are challenging the traditional unsupervised classification algorithm. In this paper we use Minimum Spanning Tree based clustering algorithm for clustering amino acid sequences. A similarity graph is defined and a cluster in that graph corresponds to connected sub graph. Cluster analysis seeks grouping of amino acid sequence in to subsets based on Euclidean distance between pairs of sequences. Our goal is to find disjoint subsets, called clusters, such that two
Recuperating Link Structure of Website using MST
"... With the explosive growth of information sources available on the World Wide Web, it has become increasingly necessary for users to utilize automated tools in find the desired information resources, and to track and analyze their usage patterns. These factors give rise to the necessity of creating s ..."
Abstract
 Add to MetaCart
With the explosive growth of information sources available on the World Wide Web, it has become increasingly necessary for users to utilize automated tools in find the desired information resources, and to track and analyze their usage patterns. These factors give rise to the necessity of creating server side and client side intelligent systems that can effectively mine for knowledge. Web mining can be broadly defined as the discovery and analysis of useful information from the World Wide Web. Our aim is to utilize the concept of using an efficient MST approach to assist easier web navigation with involvement of clustering concept. In this paper we analyze and thereby make use of implementing the same concept in it, which works on the fact of constructing a minimum spanning tree of a point set i.e. nodes and removes edges that satisfy a predefined criterion.
Meta Similarity Fine Clusters Using Dynamic Minimum Spanning Tree with SelfDetection of Best Number of Clusters
"... Clustering is a process of discovering groups of objects such that the objects of the same group are similar, and objects belonging to different groups are dissimilar. A number of clustering algorithms exist that can solve the problem of clustering, but most of them are very sensitive to their input ..."
Abstract
 Add to MetaCart
Clustering is a process of discovering groups of objects such that the objects of the same group are similar, and objects belonging to different groups are dissimilar. A number of clustering algorithms exist that can solve the problem of clustering, but most of them are very sensitive to their input parameters. Minimum Spanning Tree clustering algorithm is capable of detecting clusters with irregular boundaries. Detecting outlier in database (as unusual objects) is a big desire. In data mining detection of anomalous pattern in data is more interesting than detecting inliers. In this paper I propose a Minimum Spanning Tree based clustering algorithm for fine or pure clusters. The algorithm constructs hierarchy from top to bottom. At each hierarchical level, it optimizes the number of cluster, from which the proper hierarchical structure of underlying dataset can be found. The algorithm uses a new cluster validation criterion based on the geometric property of data partition of the data set in order to find the proper number of clusters at each level. The algorithm works in three phases. The first phase of the algorithm create rough clusters by removing outliers from data set with guaranteed intracluster similarity, where as the second phase of the algorithm removes local outliers from the rough clusters. The third phase creates dendrogram using the fine clusters as objects with guaranteed intercluster similarity. The first phase of the algorithm uses divisive approach, where as the second phase uses agglomerative approach. In this paper I used both the approaches in the algorithm to find Best Meta similarity fine clusters.
Hybrid Algorithm for Noisefree High Density Clusters with SelfDetection of Best Number of Clusters
"... Clustering is a process of discovering group of objects such that the objects of the same group are similar, and objects belonging to different groups are dissimilar. A number of clustering algorithms exist that can solve the problem of clustering, but most of them are very sensitive to their input ..."
Abstract
 Add to MetaCart
Clustering is a process of discovering group of objects such that the objects of the same group are similar, and objects belonging to different groups are dissimilar. A number of clustering algorithms exist that can solve the problem of clustering, but most of them are very sensitive to their input parameters. Minimum Spanning Tree clustering algorithm is capable of detecting clusters with irregular boundaries. A densitybased notion of clusters which is designed to discover clusters of arbitrary shape. In this paper we propose a combined approach based on Minimum Spanning Tree based clustering and Densitybased clustering for noisefree high density best number of clusters. The algorithm uses a new cluster validation criterion based on the geometric property of data partition of the data set in order to find the proper number of clusters at each level. The algorithm works in two phases. The first phase of the algorithm produces subtrees (noisefree clusters). The second phase finds high density clusters from the subtrees.
Hybrid Clustering Algorithm based on Mahalanobis Distance and MST
"... Most of the clustering algorithms are based on Euclidean distance as measure of similarity between data objects. Theses algorithms also require initial setting of parameters as a prior, for example the number of clusters. The Euclidean distance is very sensitive to scales of variables involved and i ..."
Abstract
 Add to MetaCart
Most of the clustering algorithms are based on Euclidean distance as measure of similarity between data objects. Theses algorithms also require initial setting of parameters as a prior, for example the number of clusters. The Euclidean distance is very sensitive to scales of variables involved and independent of correlated variables. To conquer these drawbacks a hybrid clustering algorithm based on Mahalanobis distance is proposed in this paper. The reason for the hybridization is to relieve the user from setting the parameters in advance. The experimental results of the proposed algorithm have been presented for both synthetic and real datasets.