Filters








24,108 Hits in 6.6 sec

Diversity-Based Pool of Models for Dealing with Recurring Concepts

Chun Wai Chiu, Leandro L. Minku
2018 2018 International Joint Conference on Neural Networks (IJCNN)  
In particular, DDM and Entropy Measure were selected as the best-performing drift detection method and diversity measure respectively, for use with our proposed approach.  ...  All strategies created a new model as representative after that drift detection.  ... 
doi:10.1109/ijcnn.2018.8489190 dblp:conf/ijcnn/ChiuM18 fatcat:lhfmkh6bhbgvfbpukrxadk5cdu

Ensemble Diversity in Evolving Data Streams [chapter]

Dariusz Brzezinski, Jerzy Stefanowski
2016 Lecture Notes in Computer Science  
While diversity of ensembles has been studied in the context of static data, it has not still received such research interest for evolving data streams.  ...  Measures highlighted as capable of depicting sudden and virtual drifts over time are used as basis for detecting changes with the Page-Hinkley test.  ...  In this paper, we investigate ensemble diversity measures not only as a means of visualizing ensemble and stream characteristics, but also as a basis for drift detection.  ... 
doi:10.1007/978-3-319-46307-0_15 fatcat:gqfkvtyufrhw3eptlmm7carroi

Ensemble based on Accuracy and Diversity Weighting for Evolving Data Streams

Yange Sun, Han Shao, Bencai Zhang
2022 ˜The œinternational Arab journal of information technology  
Based on MAD, a novel online ensemble method called Accuracy and Diversity weighted Ensemble (ADE) effectively handles concept drift in data streams.  ...  In this paper, unlike conventional data-streams oriented ensemble methods, we propose a novel Measure via both Accuracy and Diversity (MAD) instead of one of them to supervise ensemble learning.  ...  So, the ensemble classifier has a good generalization. 2) The measure of diversity In our algorithm, we adopt the Q statistic as the diversity measure.  ... 
doi:10.34028/iajit/19/1/11 fatcat:dldopmxl5zbs5dz6nucpqfspua

An Adaptive Heterogeneous Online Learning Ensemble Classifier for Nonstationary Environments

Tinofirei Museba, Fulufhelo Nelwamondo, Khmaies Ouahada, Rodolfo E. Haber
2021 Computational Intelligence and Neuroscience  
In such nonstationary environments, the underlying process generating the data stream is characterized by an intrinsic nonstationary or evolving or drifting phenomenon known as concept drift.  ...  In recent years, the prevalence of technological advances has led to an enormous and ever-increasing amount of data that are now commonly available in a streaming fashion.  ...  method (DDM) was used as a drift detection mechanism to test the drift handling capabilities.  ... 
doi:10.1155/2021/6669706 pmid:33815495 pmcid:PMC7987417 fatcat:xaw4dznl2bgzbbog7iq2v7yu3a

ADES: A New Ensemble Diversity-Based Approach for Handling Concept Drift

Tinofirei Museba, Fulufhelo Nelwamondo, Khmaies Ouahada, Yugen Yi
2021 Mobile Information Systems  
at any time and improves convergence to new concepts by exploiting different amounts of ensemble diversity.  ...  This work proposes a novel and evolving data stream classifier called Adaptive Diversified Ensemble Selection Classifier (ADES) that significantly optimizes adaptation to different types of concept drifts  ...  A change detection algorithm monitors the predictions of the low diversity ensemble and when a concept drift is detected by the measure of a drop in prediction of the subsequent data chunk, the drift detection  ... 
doi:10.1155/2021/5549300 fatcat:7djzzsolineg7duijijn2yuy64

Fast Reaction to Sudden Concept Drift in the Absence of Class Labels

Osama A. Mahdi, Eric Pardede, Nawfal Ali, Jinli Cao
2020 Applied Sciences  
A data stream can be considered as a sequence of examples that arrive continuously and are potentially unbounded, such as web page visits, sensor readings and call records.  ...  One of the serious and challenging problems that appears in a data stream is concept drift. This problem occurs when the relation between the input data and the target variable changes over time.  ...  First, we propose a drift detector based on a Diversity Measure as a Drift Detection Method (DMDDM) to detect concept drift in a semisupervised environment (DMDDM-S).  ... 
doi:10.3390/app10020606 fatcat:6syml5vzirhbpcbrv2cbdmr23u

Classification systems in dynamic environments: an overview

Felipe Azevedo Pinage, Eulanda Miranda dos Santos, João Manuel Portela da Gama
2016 Wiley Interdisciplinary Reviews Data Mining and Knowledge Discovery  
In addition, we provide the definitions of the main terms used to represent this problem in the literature, such as concept drift and novelty detection.  ...  Currently, in the literature, the most common strategies used to detect changes are based on accuracy monitoring, which relies on previous knowledge of the data in order to identify whether or not correct  ...  MAIN EVENTS IN DATA STREAMS As in data streams arrive a massive quantity of examples, it is very common that the occurrence of events become a challenge in the classification tasks.  ... 
doi:10.1002/widm.1184 fatcat:mjmtv65vmfeblakxyqwpcfuok4

Ensemble learning for data stream analysis: A survey

Bartosz Krawczyk, Leandro L. Minku, João Gama, Jerzy Stefanowski, Michał Woźniak
2017 Information Fusion  
Besides presenting a comprehensive spectrum of ensemble approaches for data streams, we also discuss advanced learning concepts such as imbalanced data streams, novelty detection, active and semisupervised  ...  This paper surveys research on ensembles for data stream classification as well as regression tasks.  ...  Section 2 focuses on the main characteristics of data streams and methods dedicated to their analysis, as well as on the type of data streams and drift detection methods.  ... 
doi:10.1016/j.inffus.2017.02.004 fatcat:rfc735znxjcwdebcbjxbyx7xki

Diversity of Ensembles for Data Stream Classification [article]

Mohamed Souhayel Abassi
2019 arXiv   pre-print
However, diversity measures have not received much research interest in evolving data streams.  ...  In this paper, we present a theoretical analysis of different diversity measures and relate them to the success of ensemble learning algorithms for streaming data.  ...  In the next section, we present a few diversity measuring techniques, but this time in a data stream setting. B.  ... 
arXiv:1902.08466v1 fatcat:pa5eqgdwt5a5libyeidvgm6qdm

Empirical Study of Impact of Various Concept Drifts in Data Stream Mining Methods

Veena Mittal, Indu Kashyap
2016 International Journal of Intelligent Systems and Applications  
As a result, the concept drifts occur very frequently in the data stream.  ...  Index Terms-Concept drift, on line learn ing, data stream mining, drift detection, ensembles.  ...  Drift Detection Method (DDM) Drift Detection Method (DDM ) [11] , as mentioned above is based on drift detection, when a drift is detected the system rebuild itself to incorporate the change in the concept  ... 
doi:10.5815/ijisa.2016.12.08 fatcat:kfdiokpojnfbzlf3soe2qltxae

A Diversity Framework for Dealing With Multiple Types of Concept Drift Based on Clustering in the Model Space

Chun Wai Chiu, Leandro L. Minku
2020 IEEE Transactions on Neural Networks and Learning Systems  
We propose to use diversity as a framework to handle multiple types of drift.  ...  Data stream applications usually suffer from multiple types of concept drift.  ...  The diversity measure used in DP is in {Entropy Measure, Q-Statistic}.  ... 
doi:10.1109/tnnls.2020.3041684 pmid:33351764 fatcat:znfm7ipvibgirmh62btrzb4tou

Design of adaptive ensemble classifier for online sentiment analysis and opinion mining

Sanjeev Kumar, Ravendra Singh, Mohammad Zubair Khan, Abdulfattah Noorwali
2021 PeerJ Computer Science  
The drift detection algorithm needs to be very sensitive to change in data distribution for detecting the maximum number of drifts in the data stream.  ...  DataStream mining is a challenging task for researchers because of the change in data distribution during classification, known as concept drift.  ...  This method performs well with standard data streams, but it has a weakness, as this method uses DDM for drift detection, which is an error-based drift detection method and cannot work well when the data  ... 
doi:10.7717/peerj-cs.660 fatcat:yojujw7tbnde5jm2e5d5vxbcpy

Recurrent Adaptive Classifier Ensemble for Handling Recurring Concept Drifts

Tinofirei Museba, Fulufhelo Nelwamondo, Khmaies Ouahada, Ayokunle Akinola, Ridha Ejbali
2021 Applied Computational Intelligence and Soft Computing  
For most real-world data streams, the concept about which data is obtained may shift from time to time, a phenomenon known as concept drift.  ...  For most real-world applications such as nonstationary time-series data, concept drift often occurs in a cyclic fashion, and previously seen concepts will reappear, which supports a unique kind of concept  ...  Input: (D 1, D 2, . . . .., D t ) the streaming data chunks E t archive of ensemble models at time step t Diversity measure : Q Statistic Drift Detection Method Detect Drift Output: F t : the generalized  ... 
doi:10.1155/2021/5533777 fatcat:tfnimgu2zjd4lbc7wimou22qpa

Aspect Oriented Concept Drift Detection in High Dimensional Data Streams

Sankara Prasanna Kumar M
2020 International Journal of Advanced Trends in Computer Science and Engineering  
In this document, aspect-oriented concept drift detection (AOCDD) is projected into high-dimension data streams.  ...  To report concept drift, the AOCDD represents the diversity of data projection for the aspects that are used to frame the record structure in the target data streams.  ...  Then take a twice as many records as a new window cw.  ... 
doi:10.30534/ijatcse/2020/111922020 fatcat:iz2hmg7d4vduvltedlqmbz7bii

Reservoir of diverse adaptive learners and stacking fast hoeffding drift detection methods for evolving data streams

Ali Pesaranghader, Herna Viktor, Eric Paquet
2018 Machine Learning  
In addition, a number of methods have been developed to detect concept drifts in these streams.  ...  We introduce the Tornado framework that implements a reservoir of diverse classifiers, together with a variety of drift detection algorithms.  ...  The authors further acknowledge funding by the Canadian Natural Sciences and Engineering Research Council (NSERC) as well as the Ontario Trillium Scholarship (OTS).  ... 
doi:10.1007/s10994-018-5719-z fatcat:lkeartys4nfo3ko7epdwaxlksy
« Previous Showing results 1 — 15 out of 24,108 results