Task-Aware Neural Architecture Search
This work addresses the challenge of automating neural network design for machine learning practitioners, though it appears incremental as it builds on existing NAS techniques.
The paper tackles the problem of neural architecture search requiring domain knowledge and limited search spaces by proposing a task-aware framework that uses a dictionary of base tasks and similarity metrics to generate adaptive search spaces, with results demonstrating efficacy through a gradient-based search algorithm that avoids full network training.
The design of handcrafted neural networks requires a lot of time and resources. Recent techniques in Neural Architecture Search (NAS) have proven to be competitive or better than traditional handcrafted design, although they require domain knowledge and have generally used limited search spaces. In this paper, we propose a novel framework for neural architecture search, utilizing a dictionary of models of base tasks and the similarity between the target task and the atoms of the dictionary; hence, generating an adaptive search space based on the base models of the dictionary. By introducing a gradient-based search algorithm, we can evaluate and discover the best architecture in the search space without fully training the networks. The experimental results show the efficacy of our proposed task-aware approach.