The internet has made the entire world interconnected in the modern day. The risk of a cyber assault has increased along with advancements in communication technologies. Today’s world runs a lot around our private data that no one can afford to be compromised. Recently, there have been data breaches in big companies like Facebook, Dominoes, etc. which had very large databases and compromised data of millions of people. Such types of threats, attacks or breaches have become very common. The advancement in IoT technology has increased the risk of a data breach even further. As a result, research on cyber security has gained a lot of attention lately. One of the instruments used to find any breaches or attacks on our internet-connected devices is the network intrusion detection system (NIDS). In this paper, we have proposed an efficient model for detecting network intrusions in any device. We have proposed a Machine learning based model and have used UGR’16 dataset for proving the efficiency of our model. The paper follows stepwise execution of the experiment. Firstly, using feature engineering and feature selection, we processed and updated the dataset, then preprocessing is done on the dataset. Then the dataset is trained for the performance, the best performing model was chosen for further hyper optimization with different techniques, the performance of the proposed model came out better than the earlier experiment done before.