Improved Automated Machine Learning from Transfer Learning
This work addresses computational efficiency in automated machine learning for researchers and practitioners, but it is incremental as it builds on existing transfer learning and neural architecture search methods.
The paper tackles the problem of neural architecture search being computationally expensive by using a task similarity measure based on Fisher Information to reduce the search space, resulting in more efficient discovery of optimal architectures without training from scratch.
In this paper, we propose a neural architecture search framework based on a similarity measure between some baseline tasks and a target task. We first define the notion of the task similarity based on the log-determinant of the Fisher Information matrix. Next, we compute the task similarity from each of the baseline tasks to the target task. By utilizing the relation between a target and a set of learned baseline tasks, the search space of architectures for the target task can be significantly reduced, making the discovery of the best candidates in the set of possible architectures tractable and efficient, in terms of GPU days. This method eliminates the requirement for training the networks from scratch for a given target task as well as introducing the bias in the initialization of the search space from the human domain.