Feature Fusion for Improved Classification: Combining Dempster-Shafer Theory and Multiple CNN Architectures
This addresses the problem of unreliable predictions in complex, real-world environments for deep learning practitioners, though it is incremental as it builds on existing pre-trained models.
The paper tackled uncertainty in deep learning by introducing a novel algorithm using Dempster-Shafer Theory to integrate multiple pre-trained models into an ensemble, achieving classification accuracy improvements of 5.4% on CIFAR-10 and 8.4% on CIFAR-100 compared to the best individual models.
Addressing uncertainty in Deep Learning (DL) is essential, as it enables the development of models that can make reliable predictions and informed decisions in complex, real-world environments where data may be incomplete or ambiguous. This paper introduces a novel algorithm leveraging Dempster-Shafer Theory (DST) to integrate multiple pre-trained models to form an ensemble capable of providing more reliable and enhanced classifications. The main steps of the proposed method include feature extraction, mass function calculation, fusion, and expected utility calculation. Several experiments have been conducted on CIFAR-10 and CIFAR-100 datasets, demonstrating superior classification accuracy of the proposed DST-based method, achieving improvements of 5.4% and 8.4%, respectively, compared to the best individual pre-trained models. Results highlight the potential of DST as a robust framework for managing uncertainties related to data when applying DL in real-world scenarios.