from Lewis L. Smith
Randomness and orderliness are of course human constructs but they have an operational reality which is important to people who are interpreting the output of electronic measurements of biological activity, forecasting the stock market et cetera. Unfortunately scientists cannot agree on what the word "random" means. Among other things there are conflicting definitions and, for a long time in securities markets, a misplaced faith in the idea that "random processes" always generate "random outputs".
Nevertheless, at any given stage of human knowledge, an objective fact is the existence of outputs of dynamic systems where the next event cannot be predicted on the basis of any previous event or combination thereof. As I suggest in my paper, where this condition is dominant in the output, we may call these "independent-event generators" [ IEG's ] rather than "random systems, in order to get away from definitional issues associated with randomness.
Unfortunately again, IGE's do not always generate outputs which "look random" by some popular criterion. Many times, IEG's generate outputs which look like they came from somewhere else.
[This is also true of many other types of dynamic systems, which is why I call them "chameleon systems".]
A classic case is that in which the output can be modeled by first-order Markov chain. As is well known, this kind of IEG produces pseudo-trends and even pseudo-cycles. In fact, while its output is stationary in the mean, it is not stationary in the variance, which increases infinitely, as the "excursions" of the data from the mean become longer and longer and more and more distant. This of course why for many gambling and investment strategies, drawdown limitions always trump regression to the mean, to the great distress of the player.
Where a run of output meets some agreed standard of randomness, I suggest calling such runs [ line segments ] "operationallyi random runs".
Cordially. ###
************** Looking for a car that's sporty, fun and fits in your budget? Read reviews on AOL Autos. (http://autos.aol.com/cars-Volkswagen-Jetta-2009/expert-review?ncid=aolaut00030000000007 )
|