Loading
Life sciences · Preprint
arXiv · September 4, 2026
Posted before peer review. The findings may change or fail to hold.
This is a preprint proposing DUO, an uncertainty-aware framework for deep imbalanced regression that models heteroscedastic predictive uncertainty and applies decoupled mean-variance optimization to improve tail sample learning. The work reports superior performance on computational benchmarks (IMDB-WIKI-DIR, AgeDB-DIR, AAV2-DIR) but has not been peer reviewed and offers no clinical or real-world validation data.
Preprint. Computational benchmarks in visual (age estimation, depth prediction) and biological (protein mutation activity prediction) domains; long-tailed regression tasks with label-scarce tail samples.. Intervention: DUO: uncertainty-aware long-tailed regression framework combining conditional Gaussian distribution modeling, decoupled mean-variance optimization, and distribution-guided contrastive learning. Compared with: Existing methods (unspecified in abstract); evaluated on benchmarks IMDB-WIKI-DIR, AgeDB-DIR, AAV2-DIR.
DUO achieves best few-shot bMAE and GM metrics on IMDB-WIKI-DIR, AgeDB-DIR, and AAV2-DIR benchmarks Method remains competitive on few-shot MAE across tested benchmarks Proposes decoupled mean-variance optimization to address gradient coupling issue in heteroscedastic negative log-likelihood under imbalanced regression
Safety was not reported in the material analysed. Check the source before drawing any conclusion about harm.
This work addresses a machine learning methodology problem rather than a clinical question. Potential applications include age estimation, depth prediction, and protein mutation activity prediction, but clinical utility and validation are not established.
This is an unrefereed methodological preprint proposing a machine learning technique for imbalanced regression; it reports benchmark comparisons but has not undergone peer review.
As stated by the source record.
Quoted from the source exactly as published.
This work addresses a machine learning methodology problem rather than a clinical question. Potential applications include age estimation, depth prediction, and protein mutation activity prediction, but clinical utility and validation are not established.
Graded across the dimensions that decide whether you should act, each from what the source actually supports. There is no single score, and where a dimension was not assessed it says so.
Deep Imbalanced Regression (DIR) is pervasive in continuous prediction tasks across diverse modalities, such as age estimation, depth prediction, and protein mutation activity prediction, where label-scarce tail samples often carry higher practical value. However, most existing methods still learn deterministic point mappings under mean squared error or its simple variants, implicitly assuming a uniform uncertainty level across all samples and thereby overlooking the instance-wise heteroscedasticity that is widespread in long-tailed data. We further point out that even heteroscedastic negative log-likelihood suffers from a gradient coupling issue, which, under DIR scenarios, weakens the learning signal of hard tail samples and leads to optimization inertia as well as tail underfitting. To address this, we propose DUO, an uncertainty-aware long-tailed regression framework. Specifically, the proposed method models the regression target as a conditional Gaussian distribution to explicitly characterize instance-level predictive uncertainty, and transforms uncertainty into a dynamic enhancement signal for tail samples through decoupled mean-variance optimization. Furthermore, we design a distribution-guided contrastive learning mechanism that adaptively constructs positive and negative pairs based on the overlap between sample distributions, thereby alleviating feature looseness and cross-label semantic entanglement. Across visual and biological DIR benchmarks, DUO achieves the best few-shot bMAE and GM on IMDB-WIKI-DIR, AgeDB-DIR, and AAV2-DIR while remaining competitive on few-shot MAE.
Taken from the source record, never inferred. Follow any of these and new work involving them reaches your briefing.