The new website looks good, and I’m looking forward to seeing it complete.
I have only recently found your HTM School you tube videos, and finished watching your last episode 14 yesterday - I found the entire series very interesting and thought provoking.
I was however intrigued by the Cortical Circuitry episode, and how you can use multiple columns to feed, and back feed information. The question I now have is how do you implement this cortical circuitry idea in code?
Specifically as an example:
- You have a stream of continuous data points of temperature measurement. You want to predict what the temperature will be over the next 5 time steps of 5 minutes.
- Someone activates an external temperature device [e.g. A heater or Air Conditioner] on a certain setting. This will run for an hour, and affect the room temperature.
How can the cortical circuitry [Distal Cell States] be combined, and the learning system learn what the previous settings impact was, given the actual future recorded data?