LGGTDec 14, 2017

Nonparametric Neural Networks

arXiv:1712.05440v122 citations
Originality Incremental advance
AI Analysis

This addresses the need for efficient network architecture search without expensive global searches, though it appears incremental in method.

The paper tackles the problem of automatically determining optimal neural network size during a single training cycle, introducing nonparametric neural networks with a novel optimization algorithm and achieving promising results.

Automatically determining the optimal size of a neural network for a given task without prior information currently requires an expensive global search and training many networks from scratch. In this paper, we address the problem of automatically finding a good network size during a single training cycle. We introduce *nonparametric neural networks*, a non-probabilistic framework for conducting optimization over all possible network sizes and prove its soundness when network growth is limited via an L_p penalty. We train networks under this framework by continuously adding new units while eliminating redundant units via an L_2 penalty. We employ a novel optimization algorithm, which we term *adaptive radial-angular gradient descent* or *AdaRad*, and obtain promising results.

Foundations

The foundational work for this paper's niche, ranked by how specifically the neighbourhood builds on it — not by global fame.

Your Notes