Back

Brain Stroke Prediction: A PSO Optimized Stacked Ensemble Machine Learning Approach with SMOTEEN Data Balancing

Jain, A.; Dubey, A. K.; Gupta, M.; Yadav, S.; Panwar, A.; Atanga, R.; Lemos, B.; Mallik, S.

2025-06-08 bioengineering
10.1101/2025.06.05.657994 bioRxiv
Show abstract

Prediction of stroke is a critical challenge in healthcare, where early intervention can significantly reduce risks and improve patient outcomes. Traditional methods often struggle with imbalanced datasets and low prediction accuracy. This paper proposes a novel approach to address these issues by combining PSO-optimized Stacked Ensemble ML Classifiers with SMOTEEN data balancing to predict strokes effectively. The technique optimizes 3 Baseline ML Classifiers(RF, SVM, XGB) using Particle Swarm Optimization (PSO) and Stacked Ensemble approach is applied that significantly improving prediction performance. The proposed methodology achieved an impressive 95.57% accuracy, outperforming conventional models such as SVM, LR, RF, KNN, XGB. Overall, the PSO-optimized stacked ensemble approach outperforms traditional machine learning techniques in stroke prediction, providing a more accurate and robust model for early detection and intervention.

Matching journals

The top 5 journals account for 50% of the predicted probability mass.

50% of probability mass above

"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.