3 ms·
Making it rank multiple attributes on a scale of 1-10 also works decent in my experience. Then one can simply k-means cluster (or similar) and evaluate the gro
by dontupvoteme 3y ago
Making it rank multiple attributes on a scale of 1-10 also works decent in my experience. Then one can simply k-means cluster (or similar) and evaluate the grouping to see how accurate its estimations are
- caesil 3y agoYes, agreed. I'm doing this as well. Works excellently for NLP classifier tasks. Funnily enough, there is a certain propensity for it to output round numbers (50, 100, etc.) so I have to ask it not to do this and provide examples ("like 27, 63, or 4"). Now that I think about it I should probably randomize those.
- dontupvoteme 3y agoInteresting, I've just been doing 1-10 (maybe i should include 0) -- Do you get the same result if you floatify the larger integers, e.g. 0.000 - 10.000?