MétaCan
Menu
Back to cohort
Record W2920267096

A context-aware machine learning-based approach

2018· article· en· W2920267096 on OpenAlex

Why this work is in the frame

A frame that forgets how it found something cannot be audited. These are the routes that admitted this work.

affAt least one author lists a Canadian institution in the pinned OpenAlex snapshot.

Bibliographic record

VenueComputer Science and Software Engineering · 2018
Typearticle
Languageen
FieldComputer Science
TopicData Stream Mining Techniques
Canadian institutionsUniversity of Waterloo
Fundersnot available
KeywordsComputer scienceMachine learningArtificial intelligenceContext (archaeology)Artificial neural networkSet (abstract data type)Context modelControl (management)
DOInot available

Abstract

fetched live from OpenAlex

It is known that training a general and versatile Machine Learning (ML)-based model is more cost-effective than training several specialized ML-models for different operating contexts. However, as the volume of training information grows, the higher the probability of producing biased results. Learning bias is a critical problem for many applications, such as those related to healthcare scenarios, environmental monitoring and air traffic control. In this paper, we compare the use of a general model that was trained using all contexts against a system that is composed of a set of specialized models that was trained for each particular operating context. For this purpose, we propose a local learning approach based on context-awareness, which involves: (i) anticipating, analyzing and representing context changes; (ii) training and finding machine learning models to maximize a given scoring function for each operating context; (iii) storing trained ML-based models and associating them with corresponding operating contexts; and (iv) deploying a system that is able to select the best-fit ML-based model at runtime based on the context. To illustrate our proposed approach, we reproduce two experiments: one that uses a neural network regression-based model to perform predictions and another one that uses an evolutionary neural network-based approach to make decisions. For each application, we compare the results of the general model, which was trained based on all contexts, against the results of our proposed approach. We show that our context-aware approach can improve results by alleviating bias with different ML tasks.

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.

Full frame distilled prediction

Teacher imitation

Not calibrated prevalence, not ground truth. Human validation pending. Learned from the 10,348 direct Codex labels and 10,348 direct Gemma labels. Candidate is the union of thresholded teacher heads; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels or direct frontier model labels.

metaresearch head score (Codex)0.001
metaresearch head score (Gemma)0.000
Version: codex-gemma-dda1882f352aValidation status: machine_predicted_unvalidated
Candidate categoriesnone
Consensus categoriesnone
DomainCandidate signal: none · Consensus signal: none
Study designCandidate signal: Simulation or modeling · Consensus signal: none
GenreCandidate signal: Methods · Consensus signal: none
Teacher disagreement score0.983
Threshold uncertainty score0.758

Codex and Gemma teacher scores by category

CategoryCodexGemma
Metaresearch0.0010.000
Meta-epidemiology (narrow)0.0000.000
Meta-epidemiology (broad)0.0000.000
Bibliometrics0.0000.001
Science and technology studies0.0000.000
Scholarly communication0.0010.001
Open science0.0020.001
Research integrity0.0000.000
Insufficient payload (model declined to judge)0.0000.000

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.

Opus teacher head0.011
GPT teacher head0.212
Teacher spread0.201 · how far apart the two teachers sit on this one work
Validation statusscore_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from it