.png){kind=icon}
.png){kind=icon}
Return
A hybrid quantum-classical neural network framework for genomic sequence classification
.png){kind=icon}
PRE
AI
DOI：10.1016/j.neucom.2026.133914.png)
.png){kind=icon}
Original request for help.png){kind=icon}
Share.png){kind=icon}
Save
Abstract
En 中文
• Establishing a hybrid framework that combines classical preprocessing with QNN for effective genomic sequence classification. • Proposing a custom feature map using , , , and gates with data reuploading. This captures both nearby and distant dependencies within sequences. • Validating the performance of the QNN-GSC on four genomic benchmark datasets. • The experimental results reveal that QNN-GSC demonstrates robust performance by achieving macro-average AUC value of 0.757, average F1-score of 0.711, and average G-mean value of 0.665. • Comprehensive quantum characterization of the proposed feature map through expressibility ( ), entanglement capability ( ) and noise robustness analysis under four NISQ noise channels (Bit Flip, Phase Flip, Amplitude Damping, Depolarizing Noise).
Keywords:
Quantum neural network
Genomic sequence classification
Hybrid framework
Feature map
NISQ noise robustness