Accuracy Improvement of Flooded Area Detection from Satellite Images using Novel K-Nearest Neighbors in Comparison with Support Vector Machine

In this study, we examine the outcomes of employing Novel K-nearest Neighbors and Support Vector Machines to assess the accuracy of recognizing flooded areas. The data set that was used in this inquiry had a total of 118 samples. These samples were divided into two sets: a training set that contained 70 samples, and a testing set that contained 48 samples. An alpha value of 0.05 and a G power of 80% were used in the calculation to determine the total number of samples that were necessary for the investigation. The accuracy of the new K-nearest neighbor algorithm is 80.26 percent, whereas the accuracy of the support vector machine is 75.48%, with a significance of 0.000078 (p<0.05). The newly created K-nearest neighbor algorithm performs much better than the Support vector machine when it comes to the task of finding areas that have been flooded.