Amazon Kinesis Data Analytics now supports the real time hotspot detection that enables the user to automatically expose regions of high density in your data. For example like surging rideshare request in a particular area that will display a popular event, a high concentration of vehicles on the highway specifying traffic bottlenecks or higher sales of products within a category showing similar features. You can gain actionable insights and react to changing business and customer requirement promptly with such detecting hotspots.
You have to first call the Kinesis Data Analytics hotspot function from the kinesis application. The Hotspot function will automatically train and create appropriate machine learning model to detect subsection of the data streams that require attention. It will report and identify one or more bounding boxes of such subsection in real-time. Kinesis Data Analytics hotspot is not supervised so it means that the function doesn’t need you to label the data for model training. The model behind the scenes to adapt the alteration in the data stream of the Kinesis Data Analytics is automatically updated. Click here to learn more about the sample code and hotspot visualization and also see Detecting Hotspots on a stream in the Kinesis Data Analytics developer Guide. You can easily process the data streams in Kinesis Data Analytics in real time with standard SQL without any necessity to learn new processing frameworks or programming languages.
No comments:
Post a Comment