Abstract:Uncertain changes in data streams present challenges for machine learning models to dynamically adapt and uphold performance in real-time. Particularly, classification boundary change, also known as real concept drift, is the major cause of classification performance deterioration. However, accurately detecting real concept drift remains challenging because the theoretical foundations of existing drift detection methods - two-sample distribution tests and monitoring classification error rate, both suffer from inherent limitations such as the inability to distinguish virtual drift (changes not affecting the classification boundary, will introduce unnecessary model maintenance), limited statistical power, or high computational cost. Furthermore, no existing detection method can provide information on the trend of the drift, which could be invaluable for model maintenance. This work presents a novel real concept drift detection method based on Neighbor-Searching Discrepancy, a new statistic that measures the classification boundary difference between two samples. The proposed method is able to detect real concept drift with high accuracy while ignoring virtual drift. It can also indicate the direction of the classification boundary change by identifying the invasion or retreat of a certain class, which is also an indicator of separability change between classes. A comprehensive evaluation of 11 experiments is conducted, including empirical verification of the proposed theory using artificial datasets, and experimental comparisons with commonly used drift handling methods on real-world datasets. The results show that the proposed theory is robust against a range of distributions and dimensions, and the drift detection method outperforms state-of-the-art alternative methods.

A Drift Detection Method for Industrial Images Based on a Defect Segmentation Model

A Neighbor-Searching Discrepancy-based Drift Detection Scheme for Learning Evolving Data

Multiscale Drift Detection Test to Enable Fast Learning in Nonstationary Environments

Online Detection of Concept Drift in Visual Tracking.

Concept Drift Region Identification Via Competence-Based Discrepancy Distribution Estimation

MSIDetector: Detecting Multi-Scenario Industrial Defects Using an Adapted Visual Foundation Model and Dual Thresholding Discriminator

A Multiscale Concept Drift Detection Method for Learning from Data Streams

Multi-window Drift Detection Method Based on Integrating Outlier Identification and Input/Output Space Information with Its Application

Efficient Data-Sketches and Fine-Tuning for Early Detection of Distributional Drift in Medical Imaging

Time to Retrain? Detecting Concept Drifts in Machine Learning Systems

Unveiling dynamics changes: Singular spectrum analysis-based method for detecting concept drift in industrial data streams

Learn to Adapt: Robust Drift Detection in Security Domain

ReCDA: Concept Drift Adaptation with Representation Enhancement for Network Intrusion Detection

CADM: Confusion Model-based Detection Method for Real-drift in Chunk Data Stream

Diagnosing Concept Drift with Visual Analytics

Few-shot Semantic Segmentation for Industrial Defect Recognition

EdgeFD: An Edge-Friendly Drift-Aware Fault Diagnosis System for Industrial IoT

Online Drift Detection with Maximum Concept Discrepancy

A Method for Segmentation of Surface Defects in Non-flat Area Based on Deep Learning

Detecting and rationalizing concept drift: A feature-level approach for understanding cause–effect relationships in dynamic environments

Efficient and accurate semi-supervised semantic segmentation for industrial surface defects