Ning Wanjun / Mathematics / Faculty Mentor: Li Wang
Effectively integrating multi-view data is essential for many machine learning applications, yet challenges remain in preserving structural information while performing feature selection and regularization. To address this, we propose a graph-based multi-view learning framework that jointly optimizes structure preservation, feature selection, and regularization to enhance data representation and improve model performance.
Our method leverages graph structures to capture complex relationships across multiple views, ensuring that shared structures are retained while identifying the most informative features. By incorporating a unified optimization strategy, our approach balances structural integrity with sparsity constraints, leading to more interpretable and robust representations.
Experiments on various multi-view datasets demonstrate that our framework effectively reconstructs the intrinsic data structure and outperforms state-of-the-art methods in both clustering and classification tasks. Visualization results further validate its ability to reveal meaningful patterns, highlighting its potential for a wide range of applications in machine learning and data analysis.
Leave a Reply