Ensemble based Classification Techniques for Concept Drifting in Continuous Data Stream: A Survey
Journal Title: International Journal for Research in Applied Science and Engineering Technology (IJRASET) - Year 2016, Vol 4, Issue 12
Abstract
Data Stream Mining is a process of extracting and analyzing the hidden, predictive, knowledge based information from the rapid, fast moving and raw data streams. The technical areas of data stream mining process includes Classification, Clustering, Decision Tree, Association Rule Mining, Temporal Data Mining, Time Series Analysis, Spatial Mining, Web Mining etc. From these technical areas, Stream data classification suffered from a problem of infinite length, concept evaluation, feature evaluation and concept drift. The most challenging problem of data stream is concept-drift which refers to the deviation of data stream from one state to another unpredictable state over time. For example, vital signals of human body like ECG (Electrocardiogram), EEG (Electroencephalogram), and BP (Blood Pressure) etc. are continuous in nature and abruptly changing hence there is a need to apply an efficient real-time data stream mining techniques for taking intelligent health care decisions. In order to address concept drift evolved in these continuous data stream, a classification model must endlessly adapt itself to the most recent concept. Hence, this paper gives the overview of various ensemble based classification algorithm techniques in the field of data stream mining and explores the future directions.
Authors and Affiliations
Girish B Umaratkar, Jaykumar S Karniwar
Keywords
Related Articles
- EP ID EP22956
- DOI -
- Views 231
- Downloads 4
Review Study of Leach Protocol
WSN is a wireless Sensor Network which is a power constrained system. In WSN nodes have limited power batteries which depend upon energy efficiency. In WSN there is major issue of energy efficiency. Energy can be consum...
A Systolic Hardware Architecture of Montgomery Modular Multiplication for Public Key Cryptosystems
The Montgomery modular multiplication is mostly used in the field public-key cryptosystems. This work presents how to relax the data dependency in conventional word-based algorithms to increase the possibility of reusin...
Noval Method for Loss Allocation in Radial Distribution Systems
this paper presents an exact method for real power loss allocation to consumers connected to radial distribution networks in a deregulated environment. The proposed method has the advantage that no assumptions are made...
Performance measures of M(x) / G / 1 Queue with Balking
Queuing systems with batch arrivals / or batch services are common in a number of real situations. Baha (1986) has studied a batch arrival queue with server vacations in which the server takes a sequence of identically...
Generation Of Electricity By Means Of Footsteps: A Review
In this review paper I intend to discuss how piezoelectric effect enables us to convert the kinetic energy produced by the human footsteps to electrical energy that can be used for various applications. The energy produ...