A Soft-Voting Ensemble achieves the best behavioral-dataset attrition prediction (CV F1 = 0.737, test recall = 0.769)¶
Claim · All claims
Evidence · 1 study · associationalr2·q2
Currency · newest study 2026 · not yet checked against its sources · not reviewed by a person · recheck by 2029: tool and model accuracy changes quickly
Subclaims¶
Reading the evidence codes
q is the design tier, i the impact, n the sample, and the kind with its r
the rigour judged within that kind: what each code means.
This claim: no standardized effect is recorded in observations/ for it yet, so no benchmark reading is possible.
q2 i? On the balanced Codeforces dataset, the Soft-Voting Ensemble led nine classifiers with cross-validation F1 of 0.737 and held-out recall of 0.769. → Azuad Islam Ruhan 2026
Evidence¶
Azuad Islam Ruhan 2026¶
Azuad Islam Ruhan, Golam Mostofa Naeem, Rakibul Islam Rafi, Sherin Afrin Mim, Nazira Bani Opi, Dewan Fahad Chowdhury, and Md. Rejaul Korim Sadi. (2026). Predicting Student Attrition in Competitive Programming: A Large-Scale Study Integrating Survey Insights and Global Behavioral Logs. https://arxiv.org/abs/2608.28618 doi:10.48550/arxiv.2608.28618
q2 · i? · associational · r2
Benchmark of nine classifiers under 5-fold stratified cross-validation on the balanced behavioral dataset (n= 1,816) with an 80/20 stratified split and F1 as the primary metric. SVM (0.732) and Random Forest (0.731) performed comparably; all models fell between 0.621 and 0.737.
"The Soft-Voting Ensemble achieved the highest cross-validation F1-score of 0.737 and test recall of 0.769, correctly identifying approximately 77% of true attrition cases in the held-out set."
Discussion¶
Related Claims¶
- True attrition in competitive programming is preceded by an 83.71% reduction in monthly contest participation — related
- Engineered practice-balance features outperform raw activity counts: the Intensity Ratio ranks third among Codeforces predictors, and upsolving habit dominates the survey model — related
- A stacking ensemble of LG, SVM, NB, KNN and XGB achieved the best fine-tuning F1-score (0.88), outperforming individual models such as SVM (0.86) — a narrower finding that bears on this claim
- Rank-averaged ensembling stabilized predictions but did not outperform the single base model — related
Page metadata
| Field | Value |
|---|---|
type |
claim |
id |
soft-voting-ensemble-behavioral-attrition-f1 |
title |
A Soft-Voting Ensemble achieves the best behavioral-dataset attrition prediction (CV F1 = 0.737, test recall = 0.769) |
description |
A Soft-Voting Ensemble achieves the best behavioral-dataset attrition prediction (CV F1 = 0.737, test recall = 0.769) |
status |
review |
evidence_strength |
moderate |
newest_evidence |
2026 |
recheck_by |
2029 |
generated |
by process:wiki-ingest on 2026-10-11 |
trust |
unverified — no one has confirmed this page's content |
sources |
1 — azuad-islam-ruhan-2026 |
Frontmatter is the page's machine-readable record — see the Schema & Guide.