Diabetic detection from tongue image using segmentation and analysis compared with K-NN classifier

Machine learning technologies are gaining traction in the medical field thanks to their impressive results in illness prediction and diagnosis. In this study, we evaluate K-nearest neighbours against the Random forest method. Procedures and Supplies: The Novel Diabetic Disease dataset on Kaggle included a grand total of twenty samples. The power of a sample G is determined using Clincalc. The three parts are enrollment ratio, alpha, and power. Both the training and test datasets contain two methods, with a total of twenty entries. We test the K-nearest neighbours method and the Random Forest algorithm to see which one is more accurate. The results showed that the Random Forest algorithm achieved an accuracy of 94.10%, whilst the K-nearest neighbours method only managed 91.6310 percent. There is a significant difference between the two groups since the SPSS statistical analysis yielded a p-value less than 0.05. Using the dataset, the Random forest algorithm was able to diagnose diabetic illness more accurately than the K-nearest neighbours method.