Prediction of Insufficient Accuracy for Human Activity Recognition with Limited Range of Age using K-Nearest Neighbor

Abstract

The research study is to predict that humans are using electronic gadgets recognition to communicate with the other person. The framework to classify the Innovative K-Nearest Neighbour and Naive Bayes to perform all measures. This research study used Novel K-Nearest Neighbour and Naive Bayes to perform the operations to give the best exactness of human activity recognition in various age factor. Recent study results a 95% credibility range for the mean and standard deviation, a significance level of 0.05 percent, and 47 samples were obtained from a variety of internet sources for this study. Since the Innovative K-Nearest Neighbor algorithm and the local invariant approaches have achieved 93.05% accuracy in predicting the activity analysis, this research wants to discover greater accuracy for activity prediction using the Naive Bayes algorithm in machine learning. When applied to the analysis of human activity, the Innovative K-Nearest Neighbor algorithm achieved 90.12% accuracy, with a p-value of 0.045 (p0.05) and a 95% confidence range. According to the results of this investigation, the K-Nearest Neighbor technique executes better than the Naive Bayes approach after analyzing the effect of age on human performance.