Notice Board :





Volume III Issue IV

Author Name
O P Kolawole
Year Of Publication
2011
Volume and Issue
Volume 3 Issue 4
Abstract
Decision tree looks like a real tree, it adopts the superincumbent strategy to distribute given object into small data sets, and in these small data sets, the leaf crunodes usually connect to only one category. The decision tree is one in all the common modeling ways to classify. Firstly, this paper introduces the idea of Classification and also the technique of the choice tree. Then, this paper analyses the information of rural labor, productive expanse and also the gross output price of agriculture regarding ten cities of Bharat supported the choice tree, and adopts clump analysis technique to discretize continuous knowledge throughout the method of knowledge miming so as to perspicacity examination to the standard classification ways. Finally, generating the choice tree of our agriculture, thereby gaining the abstraction classification rules and analyzing the principles.
PaperID
2011/eusrm/03/04/1003

Author Name
H Corke
Year Of Publication
2011
Volume and Issue
Volume 3 Issue 4
Abstract
Prediction of mobile path of a mobile node is of nice importance since it reduces the football play delays incurred throughout the football play procedure. The planned system is termed prophetical quality Management theme. The authors track the movement of mobile nodes by location trailing and data processing. In location trailing, they ceaselessly monitor the movement of mobile node. The movement of mobile nodes are often foretold by the direction of movement of mobile nodes and former path history which might be obtained by data processing techniques. the {information} mining technique is employed to go looking the trail history of mobile node and uses this information to predict the long run movement of mobile node. metric weight unit pruning theme are often used for prediction in conjunction with location trailing. This paper presents the Behavior-based quality Prediction theme to eliminate the scanning overhead incurred in IEEE 802.11 networks. this is often achieved by considering not solely location info however additionally cluster, time-of-day, and period characteristics of mobile users.
PaperID
2011/eusrm/03/04/1007

Author Name
L D Terlizzi
Year Of Publication
2011
Volume and Issue
Volume 3 Issue 4
Abstract
—Optimized Rapid Sample Protocol based on Clustering Approach along with comparative results between Rapid Sample and its optimized version are presented in this paper. In the journey of development for achieving upgraded performance, Rapid sample protocol has been designed for frequent channel variation scenario. Problem of severe energy consumption and packet loss was found in Rapid Sample Protocol. Execution of this work is segregated as, existing Rapid Sample Protocol is implemented and values of the considered parameters are recorded, in next level using clustering concept optimization of Rapid Sample Protocol has been done and in final stage comparative results between initial two levels provided by plotting various graphs for individual parameters like Energy consumption Packet loss, Throughput and Packet Delivery Ratio.
PaperID
2011/eusrm/03/04/1012

Author Name
Kwach J. K.
Year Of Publication
2011
Volume and Issue
Volume 3 Issue 4
Abstract
Web mining concern with the extraction of knowledge or information from the world wide web . One of the problem in web mining is to predict user’s next web page request. Several techniques for this prediction purpose are employed and some of these are rough set clustering , support vector machine, fuzzy logic, neural network etc [1][2][3][4]. Besides these techniques Markov models are used in web page request prediction. In this paper we describe about new proposed model which is the combination of markov model and nearest neighbor algorithm named as Hybrid Markov Model(HMM).
PaperID
2011/eusrm/03/04/1017