Network API

Hi Team, thanks for providing such an active and helpful forum. I have really enjoyed knowing about HTM for the last two months.

I am trying to use the Network API, to create a custom network. I was going through API docs. It seems to me that we always need to link the sensor component to the classifier component. Why do we need to make that link? what kind of learning happens in this link? (The code block which is related to my question given below.)

Code block taken from API:

def createSensorToClassifierLinks(network, sensorRegionName,
                                  classifierRegionName):
    """Create required links from a sensor region to a classifier region."""
    network.link(sensorRegionName, classifierRegionName, "UniformLink", "",
               srcOutput="bucketIdxOut", destInput="bucketIdxIn")
    network.link(sensorRegionName, classifierRegionName, "UniformLink", "",
               srcOutput="actValueOut", destInput="actValueIn")
    network.link(sensorRegionName, classifierRegionName, "UniformLink", "",
               srcOutput="categoryOut", destInput="categoryIn")
1 Like

No learning happens in a link. A link only defines how data flows from one node in a network to another node in a network. It draws the lines of communication between the algorithms like SP and TM. The algorithms do the learning, the links are just an architectural detail that abstracts away the connection between the algorithms.