These are chat archives for FreeCodeCamp/DataScience
discussion on how we can use statistical methods to measure and improve the efficacy of http://freeCodeCamp.com
@ridaamine the graph is not determining parameters, but trends. In fact, not really that it is not telling you which parameters: you have to understand the GEOMETRY of the oscillation and try to answer it. YOU have to find the right answer to THAT trend, adding the right formula that bring that trend to a random one.
Actually, it could be an iterative process: after adding some solution for THAT specific trend you showed me, you might have to evaluate for other different lags and see if no other periodic oscillations appears (NOTE: can be NON-periodic too! Much harder to deal with). If so, you must add another terms to your formula. You have to repeat the process until you have get rid of all "oscillations" so what you get from those charts is always a random trend (there should be a way to evaluate that randomness, right now I don't remember).
Hope this helps.
satenndrra sends brownie points to @erictleung :sparkles: :thumbsup: :sparkles: