AI Paper Atlas

Research map / Label-efficient and robust learning / Recognition and long-tail

Extrapolated Markov Chain Oversampling Method for Imbalanced Text Classification

Aleksi Avela, Pauliina Ilmonen

JMLR 2026 · Label-efficient and robust learning › Recognition and long-tail · approach: Group-robust classification · 0 citations (OpenAlex)

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
add oversampling with EMCO gamma=1 to SVMLinear SVM · Reuters full text, very low frequency, 20% ratiobalanced accuracy: .656 -> .815Improved
add majority-class transitions (gamma=1 instead of plain MCO gamma=0) (ablation)EMCO · Reuters full text, very low frequency, 20% ratiobalanced accuracy: .730 -> .815Improved
add majority-class transitions (gamma=1 instead of MCO) (ablation)EMCO · Reuters titles, very low frequency, 10% ratiobalanced accuracy: .679 -> .736Improved
use EMCO instead of EDALinear SVM · HuffPost full text, very low frequency, 20% ratiobalanced accuracy: .613 -> .722Improved
use EMCO instead of EDALinear SVM · Reuters titles, low frequency, 10% ratio, HuffPost titles*balanced accuracy: .721 -> .695Mixed
use EMCO gamma=1 instead of SVM without oversamplingLinear SVM · Reuters titles, very low frequency, 10% ratioF1-score: .300 -> .277Hurt
use EMCO gamma=1 instead of SVM without oversamplingLinear SVM · Reuters full text, very low frequency, 10% ratioF2-score: .333 -> .547Improved
lower majority weight gamma from 1 to 0.1 (ablation)EMCO · Reuters titles, 20% ratio synthetic vocabularysynthetic words added: 465.68 -> 435.85 words; recall 0.57 -> 0.55Mixed
add EMCO oversampling to LSTMBiLSTM with fastText · Reuters titles, very low frequency, 20% ratiobalanced accuracy: .603 -> .705Improved
use EMCO instead of EDA for LSTM oversamplingBiLSTM with fastText · Reuters titles, very low frequency, 20% ratioF1-score: .294 -> .161Hurt
use EMCO synthetic vocabulary instead of EDA synonym replacementminority vocabulary expansion · Reuters titles, 20% ratiobalanced accuracy: 0.53 -> 0.69Improved
use EMCO vocabulary instead of DECOM uniform priorminority vocabulary expansion · Reuters titles, 20% ratiobalanced accuracy: 0.51 -> 0.69Improved

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy