Collaborative Concept Drift Detection for Formerly Independent Models and Features
Bibliographic record
Abstract
Unstationary data can cause a change or drift in the machine learning model’s context(i.e. understanding of information) and/or concept (i.e. relationship between context and\ntarget). Resultantly, the unaccounted effects of unstationary data is referred to as drift. Drift\ncan lead to model performance degradation, despite the lack of change from an optimally\nperforming model prior to drift’s occurrence. Many works have been proposed to detect and\nrecover from these moments of drift. Despite the curation of data from the data pipeline,\nmany of these detectors operate per model and per data stream, overlooking the shared\ndata pipeline between these models and their streams. In other words, although models\noperate independently, they exist in an ecosystem consisting of models, features, and streams\nintegrated altogether. Arguably, there are resources that once considered holistically can help\nimprove our understanding of factors related to drift.This work focuses mainly on concept drift. The contribution of this dissertation is the advancement made towards adaptive, global drift detection with respect to retraining costs.This will go over: i) creating model associations based on shared features using a method traditionally used for recommendation systems, ii) relating the cost of recovery after retraining\nfrom concept drift with respect to performance and iii) creating an adaptive and aggregated\napproach to detecting drift.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame machine prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.009 | 0.028 |
| Meta-epidemiology (narrow) | 0.002 | 0.001 |
| Meta-epidemiology (broad) | 0.002 | 0.002 |
| Bibliometrics | 0.003 | 0.002 |
| Science and technology studies | 0.001 | 0.001 |
| Scholarly communication | 0.004 | 0.005 |
| Open science | 0.003 | 0.004 |
| Research integrity | 0.002 | 0.004 |
| Insufficient payload (model declined to judge) | 0.002 | 0.002 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".