Loading...
Search for:
gossip-based-protocol
0.003 seconds
An Algorithm for Frequent Pattern Mining in Distributed Data Mining
, M.Sc. Thesis Sharif University of Technology ; Mirian Hossinabadi, Hassan (Supervisor)
Abstract
Distributed computation has an important role in data mining. Data mining, usually requires a lot of resources both in terms of storage space and computing time. To reach scalability, some mechanisms for the distribution of work is required. Moreover, in many systems today, data is distributed in several databases. In this case, due to traffic overload in network, transferring all data to a central server and applying process centrally is inefficient and is subject to privacy issues. Distributed data mining provides some techniques for efficient mining in decentralized mode. Identifying frequent itemsets is one of the important branches of data mining. Frequent patterns are set of items,...
Clustering in Peer-to-Peer Networks
, M.Sc. Thesis Sharif University of Technology ; Habibi, Jafar (Supervisor) ; Mirian Hosseinabadi, Hassan (Supervisor)
Abstract
Identifying clusters is an important aspect of analyzing large data sets. In many popular applications like peer-to-peer systems, large amounts of data, is distributed among multiple dispersed data sources. In such decentralized platforms, transmission of data to a central server is infeasible due to processing, storage and transmission costs. Furthermore, the data rarely remains static and frequent data changes, quickly out dates previously extracted clustering models. In this thesis, we propose a fully decentralized partition-based clustering algorithm which is capable of clustering dynamic and distributed data sets, without requiring a central control or message flooding. In this...
Association Rules Mining in Distributed and Dynamic Databases
, M.Sc. Thesis Sharif University of Technology ; Habibi, Jafar (Supervisor) ; Mirian Hosseinabadi, Hassan (Supervisor)
Abstract
Classical methods of data mining assume that data is centric, is in memory, and is static, although in reality, most of the systems have a lot of data in distributed and dynamic environments or databases. So, classical algorithms of data mining in such environments lose memory and computation resources. In this case, transferring the whole data to a central server and applying the process centrally is inefficient and is subject to privacy issues. Distributed data mining techniques try to address these problems. Mining association rules is one of the important data mining strategies which mines frequent itemsets, correlation, or random structures among itemsets in transactional databases....
Distributed Computations in Next Generation Networks
, Ph.D. Dissertation Sharif University of Technology ; Golestani, Jamaloddin (Supervisor)
Abstract
There has been a sudden emergence of next generation networks in the past decade where the primary purposes are data ggregation/mining, distributed information and signal processing, and environmental control and monitoring. The distributed algorithms operating in such networks, should have simple structure and be robust against node failures or network dynamics.Extensive studies on designing and analyzing these algorithms have resulted in introducing different models of distributed systems with similar properties such as gossip algorithms, population protocols, and cellular automata-based systems. In this dissertation, we take first steps toward understanding the computational power of...
Mining distributed frequent itemsets using a gossip based protocol
, Article Proceedings - IEEE 9th International Conference on Ubiquitous Intelligence and Computing and IEEE 9th International Conference on Autonomic and Trusted Computing, UIC-ATC 2012 ; 2012 , Pages 780-785 ; Mirian Hosseinabadi, S. H ; Mashayekhi, H ; Habibi, J ; Sharif University of Technology
2012
Abstract
Recently, there has been a growing attention in frequent itemset mining in distributed systems. In this paper, we present an algorithm to extract frequent itemsets from large distributed datasets. Our algorithm uses gossip as the communication mechanism and does not rely on any central node. In gossip based communication, nodes repeatedly select other random nodes in the system, and exchange information with them. Our algorithm proceeds in rounds and provides all nodes with the required support counts of itemsets, such that each node is able to extract the global frequent itemsets. For local iteration and generation of candidate itemsets, a trie data structure is used, which facilitates the...