Gnas

Abstract

A key problem in deep multi-attribute learning is to effectively discover the inter-attribute correlation structures. Typically, the conventional deep multi-attribute learning approaches follow the pipeline of manually designing the network architectures based on task-specific expertise prior knowledge and careful network tunings, leading to the inflexibility for various complicated scenarios in practice. Motivated by addressing this problem, we propose an efficient greedy neural architecture search approach (GNAS) to automatically discover the optimal tree-like deep architecture for multi-attribute learning. In a greedy manner, GNAS divides the optimization of global architecture into the optimizations of individual connections step by step. By iteratively updating the local architectures, the global tree-like architecture gets converged where the bottom layers are shared across relevant attributes and the branches in top layers more encode attribute-specific features. Experiments on three benchmark multi-attribute datasets show the effectiveness and compactness of neural architectures derived by GNAS, and also demonstrate the efficiency of GNAS in searching neural architectures. KEYWORDSmulti-task learning, multi-attribute analysis, neural architecture search, greedy algorithm ACM Reference Format:

Exploring the Advancements and Challenges of Automated Machine Learning

“…The Greedy Algorithm is a method that solves optimization problems by making the most suitable decision at each step in hopes of reaching the ideal global solution. It builds a solution by choosing the best option and feature [14] at each step, until a complete solution is found.For instance, in the Neural Architecture Search (NAS) problem, determining the architecture for each layer is required, and a greedy search is utilized for [15] multi-attribute learning problems.It can be applied to various problems and is simple to implement, but doesn't guarantee the optimal solution and may lead to suboptimal ones.…”

Deep visual tracking for fiducial markers: a pilot study in telerehabilitation

“…An important task when designing DNN is the choice of the network topology, which generaly depends on previous domain knowledge and expertise. Recently, the development of methods which minimize human interference (and therefore, the need for previous knowledge) has been discussed by researchers and practitioners [Huang et al 2018]. Currently used approaches include random search, grid search and transfer learning.…”