Structure-Based Virtual Sample Generation Using Average-Linkage Clustering for Small Dataset Problems
Year: 2026
Type: article
Abstract: Small datasets are often challenging due to their limited sample size. This research introduces a novel solution to these problems: average linkage virtual sample generation (ALVSG). ALVSG leverages the underlying data structure to create virtual samples, which can be used to augment the original data set. The ALVSG process consists of two steps. First, an average-linkage clustering technique is applied to the data set to create a dendrogram. The dendrogram represents the hierarchical structure of the data set, with each merging operation regarded as a linkage. Next, the linkages are combined into an average-based data set, which serves as a new representation of the data set. The second step in the ALVSG process involves generating virtual samples using the average-based data set. The research project generates a set of 100 virtual samples by uniformly distributing them within the provided boundary. These virtual samples are then added to the original data set, creating a more extensive data set with improved generalization performance. The efficacy of the ALVSG approach is validated through resampling experiments and t-tests conducted on two small real-world data sets. The experiments are conducted on three forecasting models, the support vector machine for regression (SVR), deep learning model (DL) and XGBoost. The results show that the ALVSG approach outperforms the baseline methods in terms of mean square error (MSE), root mean square error (RMSE), and mean absolute error (MAE).
Institution: National Taiwan University of Science and Technology
Language: English
Subfield: Artificial Intelligence
Field: Computer Science
Domain: Physical Sciences
Open Access status: diamond
OpenAlex ID: W7119484606