Dynamic Ensemble Member Selection for Data Stream Classification

Ensemble methods are widely recognized for their effectiveness in data stream classification. This paper introduces Dynamic Ensemble Member Selection (DEMS), a novel framework that dynamically selects a subset of classifiers from an ensemble for each individual prediction. DEMS ranks base learners based on estimated accuracy and predictive margin, using only the top-K members for prediction, where K is optimized in a self-adaptive manner. The proposed method significantly enhances predictive performance across various state-of-the-art ensemble algorithms, including Streaming Random Patches, Adaptive Random Forest, and Online Smooth Boost. Experimental results demonstrate that DEMS consistently improves classification accuracy while maintaining a minimal runtime overhead of just 11.66% compared to the original methods. This work highlights the potential of DEMS in adapting to concept drift and optimizing ensemble diversity, offering a practical solution for real-time data stream classification.

• Publication year

  2025

• Venue

  International Conference on Information and Knowledge Management

• Publication date

  2025-11-10

• Fields of study

  Computer Science

• Identifiers
  DOI 10.1145/3746252.3761072
• External record

  Open on Semantic Scholar

• Source metadata

  Semantic Scholar

• No claims are published for this paper.

• No concepts are published for this paper.

• Random Forest Pruning Techniques: A Recent Review

  2023cited by this paper
• Survey on Online Streaming Continual Learning

  2023cited by this paper
• A survey on machine learning for recurring concept drifting data streams

  2022cited by this paper
• Dynamic Ensemble Selection for Imbalanced Data Streams With Concept Drift

  2022cited by this paper
• SOKNL: A novel way of integrating K-nearest neighbours with adaptive random forest regression for data streams

  2022cited by this paper
• Learning from evolving data streams through ensembles of random patches

  2021cited by this paper
• Dynamically Selected Ensemble for Data Stream Classification

  2021cited by this paper
• On-the-Fly Ensemble Pruning in Evolving Data Streams

  2021cited by this paper
• Boosting Ensemble Accuracy by Revisiting Ensemble Diversity Metrics

  2021cited by this paper
• Preprocessed dynamic classifier ensemble selection for highly imbalanced drifted data streams

  2021cited by this paper
• Data Preprocessing and Dynamic Ensemble Selection for Imbalanced Data Stream Classification

  2019cited by this paper
• Streaming Random Patches for Evolving Data Stream Classification

  2019cited by this paper
• Classifier Selection for Highly Imbalanced Data Streams with Minority Driven Ensemble

  2019cited by this paper
• Stability of feature selection algorithm: A review

  2019cited by this paper
• Machine learning for streaming data: state of the art, challenges, and opportunities

  2019cited by this paper
• Dynamic classifier selection: Recent advances and perspectives

  2018cited by this paper
• Margin & diversity based ordering ensemble pruning

  2018cited by this paper
• The online performance estimation framework: heterogeneous ensemble learning for data streams

  2017cited by this paper
• Adaptive random forests for evolving data stream classification

  2017cited by this paper
• A Survey on Ensemble Learning for Data Stream Classification

  2017cited by this paper
• Handling Concept Drifts Using Dynamic Selection of Classifiers

  2016cited by this paper
• Characterizing concept drift

  2015cited by this paper
• Having a Blast: Meta-Learning and Heterogeneous Ensembles for Data Streams

  2015cited by this paper
• Combining block-based and online methods in learning ensembles from concept drifting data streams

  2014cited by this paper
• An Online Boosting Algorithm with Theoretical Justifications

  2012cited by this paper
• Design and Analysis of the Nomao challenge Active Learning in the Real-World

  2012cited by this paper
• Margin optimization based pruning for random forest

  2012cited by this paper
• Activity recognition using cell phone accelerometers

  2011cited by this paper
• Learning model trees from evolving data streams

  2010cited by this paper
• Adaptive Learning from Evolving Data Streams

  2009cited by this paper
• Dynamic Weighted Majority: An Ensemble Method for Drifting Concepts

  2007cited by this paper
• Learning from Time-Changing Data with Adaptive Windowing

  2007cited by this paper
• Classifier selection for majority voting

  2005cited by this paper
• Online bagging and boosting

  2005cited by this paper
• Stress-Testing Hoeffding Trees

  2005cited by this paper
• The problem of concept drift: definitions and related work

  2004cited by this paper
• Lossless Online Bayesian Bagging

  2004cited by this paper
• Smooth boosting and learning with malicious noise

  2003cited by this paper
• Measures of Diversity in Classifier Ensembles and Their Relationship with the Ensemble Accuracy

  2003cited by this paper
• A Framework for Clustering Evolving Data Streams

  2003cited by this paper
• Experimental comparisons of online and batch versions of bagging and boosting

  2001cited by this paper
• Analyzing Bagging

  2001cited by this paper
• Mining time-changing data streams

  2001cited by this paper
• A streaming ensemble algorithm (SEA) for large-scale classification

  2001cited by this paper
• Smooth Boosting and Learning with Malicious Noise

  2001cited by this paper
• Dynamic Classifier Selection

  2000cited by this paper
• Mining high-speed data streams

  2000cited by this paper
• Diversity versus Quality in Classification Ensembles Based on Feature Selection

  2000cited by this paper
• SPLICE-2 Comparative Evaluation: Electricity Pricing

  1999cited by this paper
• The Random Subspace Method for Constructing Decision Forests

  1998cited by this paper
• Boosting the margin: A new explanation for the effectiveness of voting methods

  1997cited by this paper
• Bagging Predictors

  1996cited by this paper
• Database Mining: A Performance Perspective

  1993cited by this paper
• The Strength of Weak Learnability

  1989cited by this paper
• The perceptron: a probabilistic model for information storage and organization in the brain.

  1958cited by this paper
• The Use of Ranks to Avoid the Assumption of Normality Implicit in the Analysis of Variance

  1937cited by this paper
• Sur les fonctions convexes et les inégalités entre les valeurs moyennes

  1906cited by this paper
• Proceedings of the Twenty-Second International Joint Conference on Artificial Intelligence Learning Decision Rules from Data Streams

  year unknowncited by this paper

• No citing papers are available for this paper.