Streaming data is the future of data. Businesses need real-time solutions that can keep up with the accelerating pace and scale of information they are collecting. This is why one of the key features that Grok models incorporate is online learning (a technical term our marketing department has affectionately changed to “continuous learning”).
A question we get from time to time is, “How is Grok different from complex event processing systems?” Like Grok, CEP systems are all about streaming data, but Numenta and CEP systems use this data in different ways. We think Numenta and CEP are highly complementary and I wanted to take a moment to explain why.
Last month the Obama administration announced a new ten year $3B science initiative called the “Brain Activity Map Project” (BAM). The ultimate goal is to map all the connections in the brain and simultaneously record the activity of all the brain’s neurons.
I am often asked, “Is Numenta’s Hierarchical Temporal Memory a neural network?” (For those who don’t know, the Hierarchical Temporal Memory, or HTM, is the heart of Grok our streaming data product.)
We promised to explain how the Sparse Football Pool relates to the brain, the CLA (Numenta’s Cortical Learning Algorithm) and Intelligence. The CLA relies on Sparse Distributed Representations, a form of information representation where you have a bunch of 0s and 1s.