Table 6 Comparative analysis of PQD classification methods from literature and the proposed approach.
Method | Year | Data-Rep. | #Classes | Accuracy (%) | NR | RTVALMS |
|---|---|---|---|---|---|---|
DWT + PNN11 | 2017 | DWT | 16 | 95.0 | 20 dB | No |
RBF + NN12 | 2018 | Time-domain | 9 | 94.5 | 40 dB | No |
Seq2Seq + Bi-GRU21 | 2019 | Time-series | 96 | 98.0 | 20 dB | No |
ST + LightGBM6 | 2019 | ST | 15 | 95.1 | 20 dB | No |
ST + SVM7 | 2019 | ST | 34 | 93.5 | 20 dB | No |
GAF + CNN22 | 2021 | GAF | 21 | 96.4 | 20 dB | No |
CBAM-DenseNet18 | 2021 | GAF | 36 | 97.2 | 20 dB | No |
Improved FCN23 | 2022 | Time-series | 20 | 94.0 | 20 dB | No |
Proposed (RP + EfficientNet-SE) | 2025 | RP | 39 | 98.5 | 20 dB | Yes |
