Figure 3
![Performance Metrics of nine different combinations of the three feature extraction options: Method 1 (no feature extraction), Method 2 (EC based feature extraction) and Method 3 (FPCA-based feature extraction); and the three variable selection methods: LMM, DGB and JPTA. In addition, the performance of MildInt $[21], [22]$ (no feature extraction and MildInt network for classification) is included in combination with the three variable selection options for comparison. For blocks of the same color, the darker shade signifies better overall accuracy.](https://oup.silverchair-cdn.com/oup/backfile/Content_public/Journal/bib/25/4/10.1093_bib_bbae339/2/bbae339f3.jpeg?Expires=1749887331&Signature=umdxHnq6cSQmZ9bzCzCaYtuD5IYMUmjgvICB-5GiDlATm9qIOu~Wl8LzK8UCm~vLyPApPMZ0Aq6oB0C2CUVJRNNP1qYgpbhdB~iUAiB0oCv4myJhgdaNCvPvla5SjrqTP-QNuH2Zy7XZLaoSifIHCa0--XwuMKEN7uBGhNtmB2fwmiWtnSTznW2uETo78qS5CoacdW~swfneozty0SY1z7aq4TA7C64~325Dyd5V7lP86gWJxjxOOc4pr5jzvTNA20XkNRNoi0WCCzGWqqlALpROtDIilnGT0aD8AitafcPIooPBcQFaee4g4HMCKvco6bbWQ1fnBsGmUsNufVT6sA__&Key-Pair-Id=APKAIE5G5CRDK6RD3PGA)
Performance Metrics of nine different combinations of the three feature extraction options: Method 1 (no feature extraction), Method 2 (EC based feature extraction) and Method 3 (FPCA-based feature extraction); and the three variable selection methods: LMM, DGB and JPTA. In addition, the performance of MildInt |$[21], [22]$| (no feature extraction and MildInt network for classification) is included in combination with the three variable selection options for comparison. For blocks of the same color, the darker shade signifies better overall accuracy.
