1：Experimental Design and Method Selection:

The study involved growing pepper plants in a controlled climate room, infecting them with Fusarium disease and inoculating some with mycorrhizal fungi to form four groups. Spectral reflectance data was collected using a spectroradiometer, and machine learning algorithms (KNN, ANN, NB) were applied for classification after feature extraction using wavelet transformation.

2：Sample Selection and Data Sources:

Pepper plants (Capsicum annuum) were used, with four groups: F+M+ (diseased and inoculated with AMF), F+M- (diseased and uninoculated), F-M+ (healthy and inoculated), F-M- (healthy and uninoculated). Each group had 4 pots, with 5 leaves per pot, totaling 80 samples. Data was spectral reflectance values from 350 to 2500 nm at 1 nm resolution.

3：List of Experimental Equipment and Materials:

ASD FieldSpec 3 Spectroradiometer, leaf clip, white spectralon for calibration, climate room with controlled temperature (26°C), humidity (60%), and lighting. Computer with Matlab R2015b, Intel i5 processor, 8 GB RAM.

4：Experimental Procedures and Operational Workflow:

Spectral measurements were taken using the spectroradiometer calibrated with white spectralon. Data was converted to ASCII files, wavelet transformation (using Sym5, dB2, Haar) was applied for feature extraction, and statistical values (mean, standard deviation, minimum, maximum) of wavelet coefficients were calculated. Classification was performed using KNN, ANN, and NB algorithms with cross-validation (75% training, 25% testing, repeated 10 times).

5：Data Analysis Methods:

Performance was evaluated based on classification success rates. Euclidean distance was used for KNN, back-propagation for ANN, and Bayesian probability for NB. Results were analyzed using Matlab.