To face increasing threats, Intrusion Detection Systems (IDS) demand high accuracy, short response time, and a never seen agility in recognizing evolving threats. This research explores Machine Learning (ML) with Deep Learning (DL) for IDS, and proposes a model based on ensemble voting among several classifiers. We perform testing on real-world data using an unbalanced database under a parallel setting with four classification algorithms: Decision Tree (DT), Random Forest (RF), K-nearest neighbors (KNN), and Multiple Layer Perceptron (MLP). The voting ensemble classification method is used to improve the accuracy of the model and to reduce the number of false positives. We also address the issue of explainability to increase trust in anomaly-based Network-IDS.