A New Type of Neural Network for Anomaly Detection in Sensor Data
28 Jun 17
1:30 PM
-
2:20 PM
Tracks:
Emerging Technologies
We are living in an area with an ever growing number of sensors, where streaming analytics and machine learning are critical. But what does it take to automate stream processing? How do we create machine learning models that can keep up with exploding numbers of data streams? This talk covers how to leverage HTM (Hierarchical Temporal Memory) - a new type of neural network - to create applications that learn in real time, automatically and without the need for data labeling.