As part of trying to understand the confidence/probability value returned along with a prediction, I have come across a curious situation that I’m wondering if anyone can help explain.

Consider the hotgym model (using the Network API, obtained from http://nupic.docs.numenta.org/stable/quick-start/network.html), but make two simple modifications to the input data,

- Change the time increments to be 15 minutes
- Change the data to be a (noiseless) sine wave

Hence the first few rows of the input data looks like,

After a suitable run-in period, this generates the 1-step ahead prediction (upper axis) at the specified confidence level (lower axis) shown in the image below.

I don’t understand why the confidence is consistently high for predictions that are rising (from below 0.5 towards 1) and then going over the top of the sine wave, where as the confidence for the predictions that are falling (from -0.5 towards -1) and then going around the bottom of the sine wave can be low.

Why does the CLA have more difficulty making a prediction in one region than another when the symmetry of the data would indicate (at least naively) that there is no difference between the regions?