Wyniki wyszukiwania - BazTech

1

Encrypted prefix tree for pattern mining

Bhattacharyya R., Bhattacharyya B., Chaudhuri A.

Foundations of Computing and Decision Sciences

|

2009

|

Vol. 34, No. 1

3-22

EN

Data influx at large volumes is welcome for quality outcome in knowledge discovery, but it causes concern for scalability of mining algorithms. We introduce three measures for scalable mining - bit-vector coding, data-partitioning and Transaction Prefix (TP)-tree. Following encryption with bit-vector coding, transaction records are partitioned with notion of common prefixes. A TP-tree structure is devised for arranging the data parts such that multiple records share common storage. Advantage is two-fold: additional storage reduction over bit-vector coding and mining common prefixes together. These altogether improve space-time requirement in frequent pattern mining. Experiments on dense datasets show significant improvements in performance and scalability of both candidate generation and pattern-growth algorithms.

2

Self-adaptive MANET : a centralized approach

Basu S. S., Chaudhuri A.

Foundations of Computing and Decision Sciences

|

2004

|

Vol. 29, No. 4

271-286

EN

In this paper we propose a centralized adaptive movement algorithm for mobile nodes in a MANET to maintain neighbourhood topology. A MANET (Mobile ad-hoc network) operates without a fixed infrastructure and is also mobile because of the nature of the applications they are proposed for. In our approach each node is enabled with a GPS receiver. A node in the network is identified as the leader in the beginning. Each node identifies their one hop neighbours before starting movement and broadcast the information to the leader. All nodes also transmit their position information obtained through GPS receiver to the leader in a predefined periodic interval. By analyzing the positions, the leader realizes the movement trend and neighbourhood breakage possibilities within the network and directs the nodes to control their movement to ensure the topology retention. The impact of this concept is the elimination of routing algorithm during message transfer between the nodes and thus increases in message transfer rate.

3

An algorithm for detection of transient fault in any arbitrarily connected network

DasGupta B., DasGupta S., Chaudhuri A.

Foundations of Computing and Decision Sciences

|

2001

|

Vol. 26, No. 2

133-144

EN

This paper deals with fault diagnosis problem of interconnected multiprocessor systems. Our previous work dealt with diagnosis of faulty nodes as well as faulty links but that was restricted to dead-faults only. To make the fault diagnosis more realistic - an efficient approach is presented here, which can detect transient faults as well. The algorithm presented here is truly distributed and generalized in nature. The algorithm guarantees proper faulty unit detection as long as the system remains connected even in the presence of faults i.e., the number of faulty processors should not exceed the node connectivity of the network. Another attractive feature of the algorithm is that every non-faulty node can diagnose the faulty units independently without the help of any central unit i.e., the algorithm is truly parallel.