Nonlinear Cosine Neighborhood Time Series-Based Deep Learning for the Prediction and Analysis of COVID-19 in India

Table 1

Performance measurements for prediction error for different prediction methods.

Samples obtained	Prediction error (%)
	NCTL	Hybrid AI model	Linear regression forecasting model
250	8	12	16
500	9.25	14.35	19.35
750	9.45	15.25	20.15
1000	9.85	18.15	23.25
1250	11.35	20.35	24.55
1500	12.45	22.45	25.35
1750	14.35	25.45	30.15
2000	14.85	26.15	33.45
2250	16.25	30.35	35.15
2500	19.15	33.45	40.23