Abstract:The distribution of streaming data may change over time, making the knowledge previously learned by machine learning models outdated. This phenomenon is known as concept drift. It is common in many fields, such as weather forecasting, electricity prediction, and computer vision. To address this issue, concept drift adaptation is conducted by updating the parameters or adjusting the architecture of models. However, it may involve changing complex model structures and lead to slow convergence. This challenge is further exacerbated in scenarios with scarce labels, where instance labels are difficult to obtain and only a small amount of labeled data are available. The key to overcoming these challenges is to learn the good feature representations for incoming instances. Accordingly, we propose a novel concept drift adaptation method called CDAL, which incorporates diffusion learning into an adversarial network structure to polish up feature extraction under scarce labels. By simulating a graph-based diffusion process, diffusion learning enhances feature correlation among similar instances and weakens the feature correlation among different instances. On the other hand, adversarial learning enhances the feature extractor by guiding the latter to compete with a pair of specially designed classifiers. With these two techniques, CDAL can facilitate concept drift adaptation and ensure robust performance in scenarios with scarce labels. Additionally, our model exhibits strong performance despite its simple structure, enabling rapid adaptation in the presence of concept drift. We compared our method with four existing state-of-the-art methods on seven public datasets, including synthetic and real-world datasets. The results show that our method achieves superior performance across all datasets.

A Selective Transfer Learning Method for Concept Drift Adaptation.

Concept Drift Adaptation by Exploiting Historical Knowledge

Concept Drift Adaptation with Scarce Labels

Concept Drift Adaptation with Scarce Labels: A Novel Approach Based on Diffusion and Adversarial Learning

Recent Advances in Concept Drift Adaptation Methods for Deep Learning.

Concept Drift Adaptation by Multi- stream Data Knowledge Transfer

Learning under Concept Drift: A Review

Concept Drift Adaptation with Continuous Kernel Learning

A Better Algorithm for Classifying Data Streams with Concept Drifting

Incremental Learning Imbalanced Data Streams with Concept Drift: the Dynamic Updated Ensemble Algorithm

Concept Drift Adaptation by Exploiting Drift Type

Elastic gradient boosting decision tree with adaptive iterations for concept drift adaptation

Transfer learning for concept drifting data streams in heterogeneous environments

A survey on concept drift adaptation

Tracking Concept Drifting with an Online-Optimized Incremental Learning Framework

Active Fuzzy Weighting Ensemble for Dealing with Concept Drift

Selective Prototype-Based Learning on Concept-Drifting Data Streams

Concept Drift Detection and Adaptation with Weak Supervision on Streaming Unlabeled Data

Learning Concept-Drifting Data Streams with Random Ensemble Decision Trees

Drift-detection Based Incremental Ensemble for Reacting to Different Kinds of Concept Drift

ADTCD: An Adaptive Anomaly Detection Approach Toward Concept Drift in IoT