# Leakage-Controlled Structure-Preserving Augmentation and Template-Aware Evaluation for Bangla Symptom-Based Disease Prediction

DevFeed: [Leakage-Controlled Structure-Preserving Augmentation and Template-Aware Evaluation for Bangla Symptom-Based Disease Prediction](<https://devfeed.tech/articles/leakage-controlled-structure-preserving-augmentation-and-template-aware-evaluation-for-bangla-symptom-based-disease-prediction-111619.md>)

Original publisher: [Read original article](<https://arxiv.org/abs/2610.41733>)

Author: Sabah Nushra, Shovan Samanta Turzo, Moinul Hossain, Md Kishor Morol

Published: 2026-11-02T05:00:00Z

Content type: news

Language: en

Sources: [arXiv: Machine Learning](<https://devfeed.tech/sources/arxiv-machine-learning.md>)

Topics: [data augmentation](<https://devfeed.tech/topics/data-augmentation.md>), [Machine learning](<https://devfeed.tech/topics/machine-learning.md>), [Hyperparameter optimization](<https://devfeed.tech/topics/hyperparameter-optimization.md>)

Tags: [2](<https://devfeed.tech/tags/2.md>), [abstract](<https://devfeed.tech/tags/abstract.md>), [also](<https://devfeed.tech/tags/also.md>), [arxiv](<https://devfeed.tech/tags/arxiv.md>), [audit](<https://devfeed.tech/tags/audit.md>), [balancing](<https://devfeed.tech/tags/balancing.md>), [based](<https://devfeed.tech/tags/based.md>), [bayes](<https://devfeed.tech/tags/bayes.md>), [benchmark](<https://devfeed.tech/tags/benchmark.md>), [benefit](<https://devfeed.tech/tags/benefit.md>), [bernoulli](<https://devfeed.tech/tags/bernoulli.md>), [but](<https://devfeed.tech/tags/but.md>), [catboost](<https://devfeed.tech/tags/catboost.md>), [challenge](<https://devfeed.tech/tags/challenge.md>), [class](<https://devfeed.tech/tags/class.md>), [cs-cl](<https://devfeed.tech/tags/cs-cl.md>), [cs-lg](<https://devfeed.tech/tags/cs-lg.md>), [disease](<https://devfeed.tech/tags/disease.md>), [evaluation](<https://devfeed.tech/tags/evaluation.md>), [prediction](<https://devfeed.tech/tags/prediction.md>), [structure](<https://devfeed.tech/tags/structure.md>), [template](<https://devfeed.tech/tags/template.md>)

## AI overview

On a 725-profile Bangla symptom benchmark, train-only asymmetric Bernoulli perturbation substantially improved XGBoost, LightGBM, and CatBoost, while the gain for Bernoulli Naive Bayes (BNB) did not withstand Holm correction; BNB nevertheless had the highest mean accuracy at 97.50%. The study used repeated real-profile holdouts, parent-grouped inner cross-validation, and balancing controls. A Hamming-distance audit found 91.0% of standard held-out profiles were within one symptom bit of training profiles. Accuracy fell on a same-disease challenge purged of profiles within one bit, though XGBoost and LightGBM retained large augmentation gains. The findings suggest augmentation benefits vary by model family, and accuracy should be read alongside profile similarity. The supplied text is an abstract, so details of robustness and explanation evaluations are not available.

## Source excerpt

arXiv:2610.41733v1 Announce Type: new Abstract: Bangla symptom-based disease prediction is commonly evaluated on a small, imbalanced structured benchmark containing many profiles that differ by only one or two symptoms. We test whether train-only asymmetric Bernoulli perturbation improves classification using 725 unique profiles spanning 85 diseases and 172 binary symptoms. Experiments use ten repeated real-profile holdouts and parent-grouped inner cross-validation. The 2 x 2 design separates augmentation from hyperparameter tuning, with exact duplication, random oversampling, class weighting, and SMOTEN as balancing controls. For Bernoulli Naive Bayes (BNB), the gain over tuned raw training does not withstand Holm correction, although BNB attains the highest mean accuracy (97.50%). The boosted-tree models benefit substantially: XGBoost, LightGBM, and CatBoost improve by 18.10, 28.45, and 10.09 percentage points, respectively. A Hamming-distance audit places 91.0% of standard held-out profiles within one symptom bit of the training data. Accuracy decreases on a same-disease 1-bit-purged challenge; nevertheless, XGBoost and LightGBM retain large augmentation gains. We also evaluate robustness to missing and spurious symptoms and the faithfulness of BNB explanations. Overall, Bernoulli perturbation is useful for some model families, but benchmark accuracy should be interpreted together with template proximity.