Title: Enhancing multi-view ensemble learning with zig-zag pattern-based feature set partitioning

Authors: Aditya Kumar; Jainath Yadav

Addresses: Department of Computer Science, Central University of South Bihar, Gaya, Bihar, India ' Department of Computer Science, Central University of South Bihar, Gaya, Bihar, India

Abstract: In multi-view ensemble learning (MEL), several models are trained on various data partitions, and their predictions are merged using an ensemble technique. This study suggests a novel approach called 'zig-zag pattern-based feature set partitioning'. The method involves two steps: first, calculating feature correlations using Pearson's coefficient, and second, ranking features based on mean correlation and arranging them in a zig-zag pattern. The zig-zag pattern ensures diverse and balanced feature subsets for data partitioning, improving model generalisation and reducing overfitting. Experimental results on ten high-dimensional datasets show the practical significance of the suggested strategy, which show that it outperforms previous strategies in accuracy and generalisation. Our proposed approach, with a Friedman rank of 6.9, significantly outperforms other methods (ranging from 2.29 to 5.4) and the single-view performance rank of 1.5, highlighting its superior effectiveness. This approach advances MEL, offering a practical solution for improving ensemble model performance in complex data analysis tasks.

Keywords: feature set partitioning; views construction; ensemble learning; zig-zag partitioning; classification; multi-view ensemble learning.

DOI: 10.1504/IJCSE.2025.148732

International Journal of Computational Science and Engineering, 2025 Vol.28 No.5, pp.515 - 529

Received: 21 Mar 2024
Accepted: 17 Jul 2024
Published online: 22 Sep 2025 *

Full-text access for editors Full-text access for subscribers Purchase this article Comment on this article