Exploiting Boosting in Hyperdimensional Computing for Enhanced Reliability in Healthcare
This work addresses reliability issues in healthcare applications where data is limited, offering a method to enhance model robustness and precision, though it appears incremental as it combines existing boosting techniques with HDC.
The paper tackles the problem of underutilization in hyperdimensional computing (HDC) leading to overfitting and reduced reliability, particularly in healthcare, by introducing BoostHD, which applies boosting to partition hyperdimensional spaces into subspaces for an ensemble of weak learners, achieving 98.37% accuracy on the WESAD dataset and outperforming state-of-the-art methods like Random Forest and OnlineHD.
Hyperdimensional computing (HDC) enables efficient data encoding and processing in high-dimensional space, benefiting machine learning and data analysis. However, underutilization of these spaces can lead to overfitting and reduced model reliability, especially in data-limited systems a critical issue in sectors like healthcare that demand robustness and consistent performance. We introduce BoostHD, an approach that applies boosting algorithms to partition the hyperdimensional space into subspaces, creating an ensemble of weak learners. By integrating boosting with HDC, BoostHD enhances performance and reliability beyond existing HDC methods. Our analysis highlights the importance of efficient utilization of hyperdimensional spaces for improved model performance. Experiments on healthcare datasets show that BoostHD outperforms state-of-the-art methods. On the WESAD dataset, it achieved an accuracy of 98.37%, surpassing Random Forest, XGBoost, and OnlineHD. BoostHD also demonstrated superior inference efficiency and stability, maintaining high accuracy under data imbalance and noise. In person-specific evaluations, it achieved an average accuracy of 96.19%, outperforming other models. By addressing the limitations of both boosting and HDC, BoostHD expands the applicability of HDC in critical domains where reliability and precision are paramount.