Lung Cancer Identification System to Improve the Accuracy Using Novel K Nearest Neighbour in Comparison with Logistic Regression Algorithm

Abstract

The K Nearest Neighbor (KNN) algorithm is going to be compared against the logistic regression method in an effort to determine whether one has the potential to provide a lower false detection rate of lung cancer. Both the Techniques and the Materials: A total of 304 photos were taken using data from three different lung cancer datasets found on Kaggle. Group 1 is the representation of the KNN method, while Group 2 is the representation of the logistic regression technique. The G power was calculated using a significance level of 80% and an alpha value of 0.05. The first group, Group 1, and the second group, Group 2, each had 20 samples analyzed. The results showed that KNN had an accuracy of 89.56 percent, but the accuracy of the logistic regression approach was only 80.11 percent. The KNN technique reached a level of significance of $\mathbf{p}=\mathbf{.042}$ when it was applied using the logistic regression methodology. The results of this research reveal that the KNN technique is much more accurate than the Logistic Regression strategy when it comes to the detection of lung cancer in the datasets that were examined for this research.