Discriminant Analysis of Apple Origin Based on Machine Learning Algorithm

Preprocessed method	PLSDA		ELM		SVM
Preprocessed method	Training set	Testing set	Training set	Testing set	Training set	Testing set
Original	53.75	51.25	91.875	83.75	98.75	91.25
1-Der	56.875	52.5	90.625	83.75	100	91.25
2-Der	73.125	66.25	88.75	85	100	100
SNV	53.125	58.75	90	78.75	100	100
MSC	82.5	81.25	95	91.25	100	100
WT	47.5	43.75	86.875	85	98.125	88.75
SG	50.625	50	90	90	98.75	91.25
FT	53.75	51.25	88.75	83.75	98.75	91.25

Table 2. Total variance of explanation

View table

Table 2. Total variance of explanation

Principal component	Eigenvalue	Contribution rate /%	Cumulative contribution rate /%
1	201.704	80.041	80.041
2	23.113	9.172	89.213
3	14.781	5.865	95.079
4	5.890	2.337	97.416
5	2.640	1.048	98.463
6	1.652	0.655	99.119
7	0.783	0.311	99.430

Table 3. Accuracy of models using different method of selecting characteristic variableunit: %
View table
Table 3. Accuracy of models using different method of selecting characteristic variableunit: %
Method ELM SVM
Training set Testing set Training set Testing set
FS 98.75 96.25 100 100
PCA 96.875 60 99.375 70
SPA 100 92.5 98.75 93.75
CARS 100 98.75 99.375 100

Tools

Get Citation

Copy Citation Text

Lixin Zhang, Nannan Zhang, Xiao Zhang. Discriminant Analysis of Apple Origin Based on Machine Learning Algorithm[J]. Laser & Optoelectronics Progress, 2022, 59(4): 0430001

Download Citation

EndNote(RIS)BibTex Plain Text

Set citation alerts for article

Save article for my favorites