Multi-Model Fusion Based on Fractional Differential Preprocessing and PCA-SRDA for the Origin Traceability of Red Fuji Apples

Hua HUANG; Meng-di NAN; Zheng-hao LI; Qiu-ying CHEN; Ting-jie LI; Jun-xian GUO

doi:10.3964/j.issn.1000-0593(2022)10-3249-07

Journals >Spectroscopy and Spectral Analysis >Volume 42 >Issue 10 >Page 3249 > Article

Spectroscopy and Spectral Analysis
Vol. 42, Issue 10, 3249 (2022)

Multi-Model Fusion Based on Fractional Differential Preprocessing and PCA-SRDA for the Origin Traceability of Red Fuji Apples

Hua HUANG^1、1;, Meng-di NAN^1、1;, Zheng-hao LI^1、1;, Qiu-ying CHEN^1、1;, Ting-jie LI^1、1;, and Jun-xian GUO^{2、2; *;}

Author Affiliations

¹1. College of Mathematics and Physics, Xinjiang Agricultural University, Urumqi 830052, China

²2. College of Mechanical and Electrical Engineering, Xinjiang Agricultural University, Urumqi 830052, China

show less

DOI: 10.3964/j.issn.1000-0593(2022)10-3249-07 Cite this Article

Hua HUANG, Meng-di NAN, Zheng-hao LI, Qiu-ying CHEN, Ting-jie LI, Jun-xian GUO. Multi-Model Fusion Based on Fractional Differential Preprocessing and PCA-SRDA for the Origin Traceability of Red Fuji Apples[J]. Spectroscopy and Spectral Analysis, 2022, 42(10): 3249 Copy Citation Text

show less

Fig. 1. The flow chart of mult-model fusion

Download full size | View in the Article

$The near-infrared transmission spectra of apples from different regions for different fractional orders(a): 0 order; (b): 0.6 order; (c): 0.9 order; (d): 1.2 order; (e): 1.5 order; (f): 2 order$

Fig. 2. The near-infrared transmission spectra of apples from different regions for different fractional orders
(a): 0 order; (b): 0.6 order; (c): 0.9 order; (d): 1.2 order; (e): 1.5 order; (f): 2 order

Download full size | View in the Article

$The fractional differential (Step 0.6) preprocessing results of the training (a) and test (b) sets$

Fig. 3. The fractional differential (Step 0.6) preprocessing results of the training (a) and test (b) sets

Download full size | View in the Article

Fig. 4. The visualization of PCA dimension reduction results
(a): Training set; (b): Test set

Download full size | View in the Article

Fig. 5. The visualization of PCA-SRDA feature extraction results
(a): Training set; (b): Test set

Download full size | View in the Article

Fig. 6. The box diagram of Apple origin identification results (200 experiments repeated)
(a): Training set; (b): Test set

Download full size | View in the Article

产地	苹果重量/g	苹果横径/mm	苹果纵径/mm	可溶性固形物含量/Brix
新疆阿克苏	273.63±49.31	87.51±5.38	74.51±5.77	15.46±1.27
山东烟台	263.86±42.27	86.19±4.62	74.42±5.03	13.30±1.11
陕西洛川	270.23±37.58	87.40±4.44	74.31±4.91	12.73±1.15

Table 1. The routine data statistics of Red Fuji apples from three regions

View in the Article

模型	训练集精度/%	测试集精度/%
FD-LDA多模型融合集成学习	86.27±2.09	81.86±3.60
FD-SRDA多模型融合集成学习	89.37±4.47	82.16±5.48
FD-PCA-LDA多模型融合集成学习	96.75±0.49	94.33±1.68
FD-PCA-SRDA多模型融合集成学习	97.33±0.49	94.84±1.48

Table 2. The identification results of apple origin based on integrated learning model of multi-model fusion (200 experiments repeated)

Download Citation

Save the article for my favorites

Paper Information