The document describes a proposed hybrid intelligent intrusion detection system (HIIDS) that uses both machine learning and deep learning techniques. The HIIDS uses Spark MLlib classifiers like logistic regression and extreme gradient boosting in the first stage to detect anomalies. Any detected attacks then proceed to the second stage, which uses a long short-term memory autoencoder (LSTMAE) deep learning model to further classify the attacks. The HIIDS was tested on the ISCX-2012 intrusion detection dataset and achieved a high accuracy rate of up to 97.52% for detecting and classifying attacks.