In the past decade there is a terrific growth in Internet and at the same time we have seen increase in malicious attacks on government, corporate, military, financial organizations. To overcome from this attacks Intrusion Detection Systems (IDSs) are developed and accepted by many institutions to keep monitor on intrusion and additional harmful behavior. However these IDSs still have some obstacles that are low detection accuracy, False Negatives (FN) and False Positives (FP). To overcome from these problems Machine Learning (ML) techniques are used which helps in increasing the Intrusion Detection Accuracy and greatly decrease the false negative rate and false positive rate. In this paper we have considered five algorithms for evaluation namely Decision Tree (D-tree), Random Forest (RF), Gradient Boosting (GB), AdaBoost (AB), Gaussian Naïve Bayes (GNB) on UNSW-NB15 dataset. And we found that Random Forest is the best classifier based on the following matrices detection acc