CaDrift: A Time-dependent Causal Generator of Drifting Data Streams

About

This work presents Causal Drift Generator (CaDrift), a time-dependent synthetic data generator framework based on Structural Causal Models (SCMs). The framework produces a virtually infinite combination of data streams with controlled shift events and time-dependent data, making it a tool to evaluate methods under evolving data. CaDrift synthesizes various distributional and covariate shifts by drifting mapping functions of the SCM, which change underlying cause-and-effect relationships between features and the target. In addition, CaDrift models occasional perturbations by leveraging interventions in causal modeling. Experimental results show that, after distributional shift events, the accuracy of classifiers tends to drop, followed by a gradual retrieval, confirming the generator's effectiveness in simulating shifts. The framework has been made available on GitHub.

Eduardo V. L. Barboza, Jean Paul Barddal, Robert Sabourin, Rafael M. O. Cruz · 2026

Related benchmarks

Task	Dataset	Result
Data Stream Classification	CaDrift Dataset 1	Accuracy—	7
Data Stream Classification	CaDrift Dataset 2	Accuracy—	7
Data Stream Classification	CaDrift Dataset 3	Accuracy—	7
Data Stream Classification	CaDrift Dataset 4	Accuracy—	7
Data Stream Classification	CaDrift Dataset 5	Accuracy—	7
Data Stream Classification	CaDrift Dataset 6	Accuracy—	7
Data Stream Classification	CaDrift Dataset 7	Accuracy—	7
Data Stream Classification	CaDrift Dataset 8	Accuracy—	7
Data Stream Classification	Sea	Accuracy—	7
Data Stream Classification	Sine	Accuracy—	7