Abstract: Millions of individuals use the Internet daily to communicate with thousands of enterprises. Attacks on networks that have never occurred may come from abnormalities. Even though it has been researched for a long time, it is always difficult to identify and guard networks against unapproved access. Alongside these improvements, there are an increasing number of everyday Internet attacks. The proposed system uses machine learning techniques to find network anomalies. The CICIDS2017 dataset was utilised to accomplish this because of how current and diverse the attacks are. To determine accuracy, precision, recall, and f1-score, various machine learning techniques, including Naive Bayes, Random Forest, ID3, K Nearest Neighbours, and MLP, were utilised.

Keywords: CICIDS2017, Naive Bayes, Random Forest, ID3, K Nearest Neighbours, MLP.

| DOI: 10.17148/IARJSET.2023.10668