Competing Astrologers might well discuss their "datasets", ask each other how they might interpret or express something. "Machine Learning" seems to be guarded

Competing Astrologers might well discuss their "datasets", ask each other how they might interpret or express something. "Machine Learning" seems to be guarded corporate secrets, more often than not. If only to keep other in the company from realizing how little there is beyond soothsaying there. WE have more astrological data now, and its easier to manipulate; but no one (that I've seen) has yet tried to patch the two together. Surely machine learning techniques can extract as much valid insight from historical star field photos as they can from our click stream data?