Organizations: VIRTUALMAN, Fuxing Street, Taipei, 106030, Taiwan (ROC). · Graduate Institute of Biomedical Engineering and Bioinformatics, National Taiwan University, No. 1, Sec. 4, Roosevelt Road, Taipei, 10617, Taiwan (ROC). · Department of Computer Science and Information Engineering, National Taiwan University, No. 1, Sec. 4, Roosevelt Road, Taipei, 10617, Taiwan (ROC). · The Metabolomics Core Laboratory, Center of Genomic Medicine, National Taiwan University, No. 1, Sec. 1, Jen Ai Road, Taipei, 10617, Taiwan (ROC). · Drug Research Center, School of Pharmacy, College of Medicine, National Taiwan University, No. 1, Sec. 4, Roosevelt Road, Taipei, 10617, Taiwan (ROC).
Data imbalance frequently occurs in QSAR modeling, yet most studies focus on classification tasks, and few address regression tasks, as it is challenging to determine which regions of the data are sparse or dense. Data-level and algorithm-level imbalance-handling strategies can have different effects on model performance. In this study, we systematically compare various strategies across GCN and fingerprint-based machine learning models. To evaluate model performance more meaningfully, we partitioned the target region into several subregions based on both domain knowledge and data distribution. We focused on the region of greatest practical interest. Our results demonstrate that selecting models based on this interest-region metric leads to more reliable performance than using the total region metric. We also observed that different imbalance-handling strategies can have divergent effects depending on the model type. These findings highlight the importance of region-aware evaluation in regression QSAR modeling and provide practical guidance for selecting appropriate strategies for imbalanced datasets.
Figures & tables
Algorithm
Hyperparameter
Range
RF
max_features
0.1, 0.2, 0.4, 0.8, 1
max_depth
5, 10, 15, 20, 25, None
max_leaf_nodes
10, 100, 200, None
min_samples_leaf
1, 2, 5, 10, 20, 50
XGB
learning_rate
0.001, 0.01, 0.1, 0.3
gamma
0, 1, 5, 10
Table 1: Hyperparameter ranges for Random Forest (RF) and XGBoost (XGB) models.
Figure 1: Identification of the highest density region (HDR) for activity distribution using KDE
Figure 2: KDE-based reweighting
Figure 3: Comparison of Resampling Strategies: Original Distribution vs. KDE-Downsampling and SMOGN
Model
Treatment
Bin
MAE
Pearson
Spearman
GCN
Original
total
9.74 ± 0.5
0.55 ± 0.07
0.23 ± 0.02
GCN
KDE-based reweighting
total
11.87 ± 0.83
0.51 ± 0.05
0.2 ± 0.01
GCN
KDE-guided SMOGN
total
17.75 ± 1.73
0.42 ± 0.02
0.22 ± 0.03
GCN
KDE-based downsampling
total
14.25 ± 0.8
0.45 ± 0.03
0.2 ± 0.02
RFR
Original
total
9.57 ± 0.19
0.57 ± 0.02
0.26 ± 0.01
RFR
KDE-based reweighting
total
10.39 ± 0.38
0.57 ± 0.02
0.25 ± 0.01
Table 2: Test performance of different imbalance-handling strategies based on total MSE selection.
Figure 4: Relationship between training data distribution and test MAE for GCN under the downsampling strategy.
Figure 5: The impact of model selection strategy on test MAE.
Imbalanced learning is a critical challenge in machine learning, where underrepresented target values can bias models and degrade prediction performance on rare but important cases. Although extensively studied in classification, imbalanced regression remains relatively underexplored. Existing methods mainly focus on either data-level balancing, which may introduce noise and overfitting, or algorithm-level balancing, which often struggles with highly complex target distributions. To address these limitations, we propose a unified hybrid framework that integrates both data- and algorithm-level balancing strategies into a regressor-agnostic pipeline. The proposed framework consists of five stages: (1) adaptive bin partitioning to dynamically segment the target space based on local linear coherence; (2) target-conditioned representation learning using a Conditional Variational Autoencoder; (3) multistage data-level balancing through feature-space clustering and oversampling of minority clusters; (4) algorithm-level balancing using a novel Latent-Density Weighted Loss (LDWL) to emphasize rare samples in latent and target spaces; and (5) attention-based gated fusion for final regression. Experimental results on benchmark datasets demonstrate that the proposed framework consistently improves predictive performance compared to standalone regressors and existing imbalanced regression approaches.
Objective: The primary goal of this study was to systematically examine the impact of commonly used imbalance handling methods (IHMs) on predictive performance in biomedical binary classification, considering the interplay between model complexity and diverse data modalities. Material and Methods: We evaluated five representative IHMs: random undersampling (RUS), random oversampling (ROS), SMOTE, re-weighting (RW), and direct F1-score optimization (DMO), against a raw training (RAW) baseline. The evaluation encompassed three public biomedical datasets: MIMIC-III (tabular), ADE-Corpus-V2 (text), and MURA (image), spanning three common biomedical data modalities. To assess varying model complexity, we employed a range of architectures, from classical logistic regression and random forest to deep neural networks, including multilayer perceptron (MLP), BiLSTM, BERT, DenseNet, and DINOv2. Results: For simpler models such as logistic regression on tabular data, IHMs yielded no significant advantage over the RAW baseline, aligning with prior findings. However, clear benefits were observed for more complex models and unstructured data: (a) ROS and RW consistently enhanced the performance of powerful models; (b) direct F1-score optimization demonstrated utility primarily for unstructured text and image data; and (c) RUS and SMOTE consistently degraded performance and are therefore not recommended. Conclusion: The effectiveness of IHMs depends on both model complexity and data modality. Performance gains are most pronounced when leveraging appropriate IHMs, such as ROS, RW, and DMO, on high-complexity models.
Jiandong Chen, Lingjie Su, Le Peng +3
Institute for Health Informatics, University of Minnesota, Minneapolis, MN · Department of Computer Science and Engineering, University of Minnesota, Minneapolis, MN · School of Statistics, University of Minnesota, Minneapolis, MN +1
Data imbalance persists as a pervasive challenge in regression tasks, introducing bias in model performance and undermining predictive reliability. This is particularly detrimental in applications aimed at predicting rare events that fall outside of the domain of the bulk of the training data. In this study, we review the current state-of-the-art regarding sampling-based methods and cost-sensitive learning. Additionally, we propose novel approaches to mitigate model bias. To better assess the importance of data, we introduce the density-distance and density-ratio relevance functions, which effectively integrate empirical frequency of data with domain-specific preferences, offering enhanced interpretability for end-users. Furthermore, we present advanced mitigation techniques (cSMOGN and crbSMOGN), which build upon and improve existing sampling methods. In a quantitative evaluation, we benchmark state-of-the-art methods on 10 synthetic and 42 real-world datasets, using neural networks, XGBoosting trees and Random Forest models. Our analysis shows that while most strategies improve performance on rare samples, they degrade it on frequent ones. The trade-off becomes larger the more the performance on rare samples is increased. However, to reduce this effect we demonstrate that constructing an ensemble of models -- one trained with imbalance mitigation and another without -- can be used. The key findings underscore the superior performance of our novel crbSMOGN sampling technique with the density-ratio relevance function for neural networks, outperforming state-of-the-art methods.
Jelke Wibbeke, Sebastian Rohjans, Andreas Rauh
Department of Computing Science Carl von Ossietzky Universität Oldenburg · Energy Division, OFFIS - Institute of Information Technology, Oldenburg, Germany · Department for Civil Engineering, Geoinformation and Health Technology Jade University of Applied Science