Research Article
Mid-Infrared Spectroscopy for Coffee Variety Identification: Comparison of Pattern Recognition Methods
Table 2
Classification results of the models on optimal wavenumbers.
| | | Training | Test | | | 1 | 2 | 3 | 4 | Total | 1 | 2 | 3 | 4 | Total | | | Nr/Nt | Nr/Nt | Nr/Nt | Nr/Nt | Nr/Nt | Accuracy (%) | Nr/Nt | Nr/Nt | Nr/Nt | Nr/Nt | Nr/Nt | Accuracy (%) |
| | SIMCA | 20/20 | 20/20 | 17/20 | 20/20 | 77/80 | 96.25 | 10/10 | 10/10 | 8/10 | 10/10 | 38/40 | 95.00 | | SVM | 20/20 | 20/20 | 20/20 | 20/20 | 80/80 | 100.00 | 10/10 | 10/10 | 9/10 | 9/10 | 38/40 | 95.00 | | ELM | 20/20 | 20/20 | 20/20 | 20/20 | 80/80 | 100.00 | 9/10 | 10/10 | 10/10 | 10/10 | 39/40 | 97.50 | | BP | 20/20 | 20/20 | 20/20 | 20/20 | 80/80 | 100.00 | 10/10 | 10/10 | 10/10 | 10/10 | 40/40 | 100.00 | | RBF | 20/20 | 20/20 | 20/20 | 20/20 | 80/80 | 100.00 | 10/10 | 10/10 | 10/10 | 10/10 | 40/40 | 100.00 | | NN | 20/20 | 20/20 | 17/20 | 19/20 | 76/80 | 95.00 | 10/10 | 10/10 | 8/10 | 8/10 | 36/40 | 90.00 | | RF | 20/20 | 20/20 | 20/20 | 20/20 | 80/80 | 100.00 | 10/10 | 10/10 | 8/10 | 9/10 | 37/40 | 92.50 | | PLS-DA | 19/20 | 18/20 | 18/20 | 16/20 | 71/80 | 86.25 | 7/10 | 8/10 | 7/10 | 10/10 | 32/40 | 80.00 | | Naive Bayes classifier | 19/20 | 17/20 | 13/20 | 9/20 | 58/80 | 72.50 | 10/10 | 10/10 | 4/10 | 5/10 | 29/40 | 72.50 | | RVM | 20/20 | 20/20 | 20/20 | 20/20 | 80/80 | 100.00 | 10/10 | 10/10 | 9/10 | 9/10 | 38/40 | 95.00 |
|
|
a: Nr was the number of correctly classified samples; b: Nt was the total number of the samples.
|
