by Angela Guess
According to a new article out of the company, “Numenta, Inc., a leader in machine intelligence, today announced the release of HTM Studio for anomaly detection, a desktop tool that makes it easy for businesses to experiment with advanced machine intelligence algorithms to uncover anomalies in their streaming data. Simply add numeric, time-series data files to HTM Studio and with the click of a button, discover and visualize within minutes what anomalies can be found. HTM Studio is based on Numenta’s HTM (Hierarchical Temporal Memory) biologically inspired machine intelligence technology. HTM learns time-based patterns in unlabeled streaming data, making it ideal for prediction, classification and anomaly detection. Because of its ability to perform continuous, unsupervised learning, HTM can find subtle, temporal anomalies that other techniques cannot find.”
The article goes on, “HTM Studio provides an easy way for businesses interested in using Numenta’s technology for anomaly detection to undertake a proof of concept securely using their own data. If HTM Studio finds interesting anomalies, then a user can do a full implementation or deployment of HTM. ‘Detecting anomalies early in streaming data can have significant value, but it’s difficult to do. We wanted to make it easy to try HTM on your own streaming data,’ said Donna Dubinsky, CEO of Numenta. ‘With HTM Studio, no coding skills are required. Parameters are automatically set. In a matter of minutes, you can visualize anomalies in your datasets.’ For those who don’t have streaming data readily available, HTM Studio includes pre-loaded datasets for experimentation. Sample datasets include a variety of use cases: IoT sensors, preventative maintenance, tracking vehicles and network servers. The tool is available at no charge for Mac and Windows desktop systems. Go to http://numenta.com/htm-studio/ for details and to get started with HTM Studio.”
Photo credit: Numenta