Predictive PAC learnability: a paradigm for learning from exchangeable input data

Details Predictive PAC learnability: a paradigm for learning from exchangeable input data Predictive PAC learnability: a paradigm for learning from exchangeable input data

2010-06-06

arxiv.org/abs/1006.1129v2

Proc. 2010 IEEE International Conference on Granular Computing (GrC 2010), San Jose, CA, August 14-16, 2010, IEEE Computer Soc

Computer Science

Learning

5 pages, latex, a postprint correcting a typo in the main definition 4.1

Scientific paper

Exchangeable random variables form an important and well-studied generalization of i.i.d. variables, however simple examples show that no nontrivial concept or function classes are PAC learnable under general exchangeable data inputs $X_1,X_2,\ldots$. Inspired by the work of Berti and Rigo on a Glivenko--Cantelli theorem for exchangeable inputs, we propose a new paradigm, adequate for learning from exchangeable data: predictive PAC learnability. A learning rule $\mathcal L$ for a function class $\mathscr F$ is predictive PAC if for every $\e,\delta>0$ and each function $f\in {\mathscr F}$, whenever $\abs{\sigma}\geq s(\delta,\e)$, we have with confidence $1-\delta$ that the expected difference between $f(X_{n+1})$ and the image of $f\vert\sigma$ under $\mathcal L$ does not exceed $\e$ conditionally on $X_1,X_2,\ldots,X_n$. Thus, instead of learning the function $f$ as such, we are learning to a given accuracy $\e$ the predictive behaviour of $f$ at the future points $X_i(\omega)$, $i>n$ of the sample path. Using de Finetti's theorem, we show that if a universally separable function class $\mathscr F$ is distribution-free PAC learnable under i.i.d. inputs, then it is distribution-free predictive PAC learnable under exchangeable inputs, with a slightly worse sample complexity.

Affiliated with

Also associated with

No associations

Rating

Predictive PAC learnability: a paradigm for learning from exchangeable input data does not yet have a rating. At this time, there are no reviews or comments for this scientific paper.

If you have personal experience with Predictive PAC learnability: a paradigm for learning from exchangeable input data, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Predictive PAC learnability: a paradigm for learning from exchangeable input data will most certainly appreciate the feedback.

Rate now

Comments { 0 }

Profile ID: LFWR-SCP-O-682334