Machine learning features in X-Pack let you automate the task of detecting anomalies in time series data. In the third video in this tutorial series, we show you how to configure an advanced job to detect anomalies (or outliers) in a population.
Download the example from GitHub to try this out on your machine: https://github.com/elastic/examples/tree/master/Machine%20Learning/Getting%20started%20examples
If you are just getting started, watch the previous tutorials in this series to learn about single metric and multimetric jobs.
Video 1: http://www.elastic.co/videos/machine-learning-tutorial-creating-a-single-metric-job
Video 2: http://www.elastic.co/videos/machine-learning-tutorial-creating-a-multi-metric-job
You can watch this video also at the source.