**3. Results**

In this section, the results are presented and discussed. The transformer oil testing samples based on mineral oil were abstracted from a South African transformer manufacturer after laboratory analysis. The BCSVM training phase was ingested with 70% of the oil testing samples and tested using 30% of the transformer oil samples. This investigation was conducted by employing the classification learner app in the MATLAB\_R2018a software platform. Once the oil samples dataset was trained, the veracity of the distinctive SVMs was corroborated. The configuration matrix of a particular SVM that provides the highest degree of accuracy was selected and imported to predict the new testing data sample. The fault class prediction for the new data sample is carried out by utilizing Equation (7).

$$yfit = fitecocc(Tbl, \text{ }ResponseVarName) \tag{7}$$

Here,

*Tbl*¯Multiclass predictor variables in table *ResponseVarName*¯Response variables
