Not waving
Mar 2nd, 2008 by handolio
I’m lucky enough to have a regular freelance gig, explaining to the good readers of Computer Shopper ‘how it works’. The only problem is that I generally don’t know how it works, and it can be extremely hard to learn enough about a topic to write it up quickly and competently, and to do so without going softly insane.
This month, the ‘it’ in question is handwriting recognition, and it’s proved a particularly hard bastard. Most of it is, in fact, pretty simple - comparing potential matches to possibilities contained within a dictionary, a bit like predictive text - but nobody seems to know or want to talk about how the actual recognition bit works.
By way of an illustration of how tricky things can get for the technical writer, I found myself asking Wikipedia about the stochastic finite state automata model.
Well, it seemed like a good idea at the time.
In the simplest possible case (’discrete time’), a stochastic process amounts to a sequence of random variables known as a time series (for example, see Markov chain). Another basic type of a stochastic process is a random field, whose domain is a region of space, in other words, a random function whose arguments are drawn from a range of continuously changing values. One approach to stochastic processes treats them as functions of one or several deterministic arguments (’inputs’, in most cases regarded as ‘time’) whose values (’outputs’) are random variables: non-deterministic (single) quantities which have certain probability distributions. Random variables corresponding to various times (or points, in the case of random fields) may be completely different. The main requirement is that these different random quantities all have the same ‘type’. Although the random values of a stochastic process at different times may be independent random variables, in most commonly considered situations they exhibit complicated statistical correlations.
It’s late Saturday night, are the pubs still open? I understand beer.
