Probably Approximately Correct (PAC) Learning Model
Probably Approximately Correct (PAC) Learning Model
- is a framework for the mathematical analysis of machine learning
- it was proposed in 1984 by Leslie Valiant
- it aims to address the challenge of building models that can generalize well to unseen data
- it defines the probability that the available training data are large enough to attain the true error rate
PAC - Framework
In this framework, the learner receives samples and must select a generalization function (called the hypothesis) from a certain class of possible functions. The goal is that, with high probability (the "probably" part), the selected function will have low generalization error (the "approximately correct" part).
see:
- Chapter 2 - Empirical Risk Minimization (ERM)
- Chapter 3 - A Formal Learning Model - Probably Approximately Correct (PAC)
Resources
, multiple selections available,