Abstract: I have been working on anomaly detection algorithms, but find that a visual approach is also very powerful. Since I want to present in an Ignite format, this aspect of my work would fit very well with the presentation format. Anomaly Detection is at the cross-roads of Data, Algorithms and Visualization: The algorithm to be used depends on what kind of data you have, and without a good visualization, its hard to understand the meaning of the anomaly detected. For DevOps in particular, the correlation of anomalies is usually more important than a single anomaly, placing higher emphasis on good visualization. In my talk, I will share some of my efforts at detecting -- and understanding the root causes -- behind anomalies using ElasticSearch. I will share beautiful visualizations and show how crafting a good visualization contributes a good deal towards visually detecting, and understanding, anomalies in a complex system such as ElasticSearch.
Speaker: Speaker 37