Incorporating Machine Learning into Vibration Detection for Wind Turbines

Table 1

Advantages and disadvantages of KNN and SVM algorithms.

Algorithm	Advantages	Limitations
KNN	Easy to implement; Good overfitting; Robustness to noisy data; Can be used for both classification and regression	Great computation; Slow sorting speed; Needs a lot of storage space; The selection of influences the classification
SVM	High sorting accuracy; Fast sorting speed; Can handle high-dimensional features; Robust to overfitting and noise	Low efficiency for big data; Binary classifier; No physical meaning