Exponentially Weighted Moving Average Charts for Detecting Concept Drift

Abstract

Classifying streaming data requires the development of methods which are computationally efficient and able to cope with changes in the underlying distribution of the stream, a phenomenon known in the literature as concept drift. We propose a new method for detecting concept drift which uses an Exponentially Weighted Moving Average (EWMA) chart to monitor the misclassification rate of an streaming classifier. Our approach is modular and can hence be run in parallel with any underlying classifier to provide an additional layer of concept drift detection. Moreover our method is computationally efficient with overhead O(1) and works in a fully online manner with no need to store data points in memory. Unlike many existing approaches to concept drift detection, our method allows the rate of false positive detections to be controlled and kept constant over time.

Cited by in corpus (34)

• Learning under Concept Drift: A Review
• Challenges in Benchmarking Stream Learning Algorithms with Real-world Data
• Handling Adversarial Concept Drift in Streaming Data
• Reservoir of Diverse Adaptive Learners and Stacking Fast Hoeffding Drift Detection Methods for Evolving Data Streams
• How good is the Electricity benchmark for evaluating concept drift adaptation
• Learning Parameter Distributions to Detect Concept Drift in Data Streams
• Concept Drift Adaptation in Text Stream Mining Settings: A Systematic Review
• Multi-Source Transfer Learning for Non-Stationary Environments
• Distribution-Free One-Pass Learning
• SoK: Machine Learning Governance
• Detection of data drift and outliers affecting machine learning model performance over time
• IncA-DES: An incremental and adaptive dynamic ensemble selection approach using online K-d tree neighborhood search for data streams with concept drift
• Class Distribution Monitoring for Concept Drift Detection
• Modyn: Data-Centric Machine Learning Pipeline Orchestration
• OPTWIN: Drift identification with optimal sub-windows
• Monitoring Term Drift Based on Semantic Consistency in an Evolving Vector Field
• Sequential Drift Detection in Deep Learning Classifiers
• Change Detection in Multivariate Datastreams: Likelihood and Detectability Loss
• Finite-time Analysis of Globally Nonstationary Multi-Armed Bandits
• Concept Drift Monitoring and Diagnostics of Supervised Learning Models via Score Vectors
• Fast likelihood-based change point detection
• Unsupervised Behavior Change Detection in Multidimensional Data Streams for Maritime Traffic Monitoring
• Concept Drift Detection and Adaptation with Hierarchical Hypothesis Testing
• Labelling Drifts in a Fault Detection System for Wind Turbine Maintenance
• A Self-Organizing Clustering System for Unsupervised Distribution Shift Detection
• Challenges in Vessel Behavior and Anomaly Detection: From Classical Machine Learning to Deep Learning
• DeepStreamCE: A Streaming Approach to Concept Evolution Detection in Deep Neural Networks
• Gradual Drift Detection in Process Models Using Conformance Metrics
• Concept Drift and Covariate Shift Detection Ensemble with Lagged Labels
• A Practical Guide to Streaming Continual Learning
• SAFE: Spectral Evolution Analysis Feature Extraction for Non-Stationary Time Series Prediction
• Parameter Estimation in Abruptly Changing Dynamic Environments
• On the Reliable Detection of Concept Drift from Streaming Unlabeled Data
• Tackling Virtual and Real Concept Drifts: An Adaptive Gaussian Mixture Model