»Hierarchical Temporal Memory in Python«
2019-06-04, 11:45–12:25, Hall 2 (PyData)

Hierarchical Temporal Memory is a novel framework for biological and machine intelligence. It learns patterns from relatively little data and is well suited for prediction, anomaly detection, classification and ultimately sensorimotor applications.

Hierarchical temporal memory (HTM) is a biologically constrained theory (or model) of intelligence, originally described in the 2004 book On Intelligence by Jeff Hawkins with Sandra Blakeslee. HTM is based on neuroscience and the physiology and interaction of neurons in the neocortex of the mammalian (in particular, human) brain. At the core of HTM are learning algorithms that can store, learn, infer and recall high-order sequences. Unlike most other machine learning methods, HTM learns (in an unsupervised fashion) time-based patterns in unlabelled data on a continuous basis. HTM is robust to noise, and it has high capacity, meaning that it can learn multiple patterns simultaneously. When applied to computers, HTM is well suited for prediction, anomaly detection, classification and ultimately sensorimotor applications.

In this talk I will cover some basics of the mammalian neocortex, provide an introduction to Hierarchical Temporal Memory (HTM), briefly compare it to Deep Neural Nets (DNN) and show some examples of HTM in action.