Learning SMaLL Predictors
This work addresses the need for efficient machine learning models in resource-limited settings, presenting an incremental improvement with a novel method.
The paper tackles the problem of training small, resource-constrained predictors by introducing the Sparse Multiprototype Linear Learner (SMaLL) algorithm, which is inspired by learning k-DNF Boolean formulae, and demonstrates its benefits through an empirical study.
We present a new machine learning technique for training small resource-constrained predictors. Our algorithm, the Sparse Multiprototype Linear Learner (SMaLL), is inspired by the classic machine learning problem of learning $k$-DNF Boolean formulae. We present a formal derivation of our algorithm and demonstrate the benefits of our approach with a detailed empirical study.