It may be necessary to detect which applications are moving via the networks inside the internet community in order to carry out specific tasks. Internet service providers (ISPs) generally employ network traffic classification to identify the prerequisites for a connection, It thus impacts the effectiveness of the cable network at the moment. Each one of the Internet Protocol (IP) methods—bandwidth, stream, and ML—has unique advantages and disadvantages. The Machine learning approach [5–9] is well-liked these days due to its vast use across disciplines and the growing knowledge among many researchers of its methodology when specifically compared to everyone else. Results from the Naive Bayes and K-nearest algorithms are then contrasted in this study when they are applied to a networking-specific dataset obtained utilizing live stream feeds and an Ethernet software. To develop a machine learning algorithm, the pandas and numpy arrays modules, the sklearn module for Python, and other help modules are used. Our research demonstrates that the K nearest method outperforms the Support Vector Machine, Nave Bayes, and Decision Tree algorithms in terms of efficiency