Suppose I am continuously getting labeled observations, each one of which is type A or type B. I want to continuously update my belief of the probability that the next object will be of type A (or B). This should involve formulating a transition rule, something of the form: "Given that my current belief is that Pr{next object is type A} = p, and the last observation was X, my new Pr{next object A} = p' ". So the idea is continuously updating my belief based on the recent event(s), and the belief(s) I had when they happened. How do I get started? What is the simplest technique? What should I start looking into? Thanks

asked Feb 28 '12 at 22:05

Viktor%20Simjanoski's gravatar image

Viktor Simjanoski
193212529

2

You could look into hidden markov models

(Feb 28 '12 at 23:25) Leon Palafox ♦
2

Hidden Markov models are probably the way forward. A bit more simplistic are Kalman filters, which apply only to the Gaussian case.

(Feb 29 '12 at 03:23) Andreas Mueller

Is this online learning? Can you be a bit more explicit about what problem you're trying to reduce to this scenario? Is this a homework question?

(Feb 29 '12 at 08:48) Alexandre Passos ♦

Yes, it is online problem. No, it is not homework, it is industry.

(Feb 29 '12 at 10:15) Viktor Simjanoski

So have you tried any of a large family of online learning algorithms for classification? Perceptron, SGD, winnow.. ? There are fast implementations out there, most notably vowpal wabbit. If that does not solve your problem, what does?

(Feb 29 '12 at 10:19) Alexandre Passos ♦

Ultimately, the output should not be discrete. Even if I describe the problem in terms of discrete states, my current outcome should be continuous(possible through probabilities of what state I am currently in).

(Feb 29 '12 at 10:23) Viktor Simjanoski

It is easy to predict probabilities in online learning.

(Feb 29 '12 at 10:23) Alexandre Passos ♦

Can you be more precise, please? How would I use perceptron for this kind of situation, or SGD, for the matter of fact?

(Feb 29 '12 at 10:30) Viktor Simjanoski
1

It looks like your problem is more of a hidden Markov model anyway. Well, if you know the emission and transition probabilities you can just multiply them and normalize.

(Feb 29 '12 at 10:31) Alexandre Passos ♦
showing 5 of 9 show all

One Answer:

A very simple technique might be to use a n-gram model and to update the counts of the model while observing the sequence.

answered Mar 01 '12 at 14:04

Dominik's gravatar image

Dominik
1763810

Your answer
toggle preview

powered by OSQA

User submitted content is under Creative Commons: Attribution - Share Alike; Other things copyright (C) 2010, MetaOptimize LLC.