A Novel Approach for Predicting Covid 19 Recovered Cases using Nourishment by Comparing KNN over SVM

ABSTRACT

To predict the accuracy value of COVID19 recovered number of patients using Nourishment. Material and Methods: For forecasting accuracy percentage of COVID19 recovered patient health diet, Novel K Nearest Neighbour with test size (N=10) and Support Vector Machine with test size (N=10) were iterated 20 times to COVID19 recovered number of patients with g power as 80 %, threshold 0.014 and confidence interval as 95%. Sigmoid function is used in K Nearest Neighbour prediction to probability to help enhance accuracy. Results: In comparison to Support Vector Machine 66% percent Accuracy, Novel K Nearest Neighbour produced substantial results with 94 % Accuracy. Support Vector Machine and K Nearest Neighbour statistical significance is p=1.000(p<0.05) Independent sample T-test value states that the results in the study are significant. Conclusion: KNN is a straightforward and efficient algorithm for quickly building Models of machine learning. KNN predicting COVID19 Health Diet % with more accuracy. © 2022 IEEE.