Duality Between Deep Learning And Algorithm Design

This thesis introduces 'Duality Between Deep Learning And Algorithm Design'. Deep learning is a data-driven method, whereas conventional algorithm design is a knowledge-driven method. Based on their connections and complementary features, this thesis develops new methods to combine the merits of both and, in turn, improve both. Specifically: (1) Algorithm inspired deep learning model: Despite the unprecedented performance of deep learning in many computer vision and natural language processing problems, the development of deep neural networks is hindered by their black-box nature, i.e., a lack of interpretability and the need for very large training sets. To eliminate these issues, this thesis introduces the use of algorithms as modeling priors to integrate specialized knowledge of domain experts into deep learning models. From both the empirical and theoretical perspective, this thesis explains how such algorithm inspired deep learning models can achieve improved interpretability and sample efficiency. (2) Deep learning based algorithm design: In conventional algorithm design, domain experts will first develop a model to describe the mechanism behind it and then establish a mathematical algorithm to find the solution. Notwithstanding its interpretability, this model-based method is inferior in terms of its limited effective range and accuracy. This is mostly due to the simplifying assumptions of the models which often deviate from real-world problems. To address these issues, this thesis investigates the potential of deep learning based methods for discovering data-driven algorithms that adapt better to the interested problem distribution. This thesis explains how to design data-driven components to replace some fixed procedures in traditional algorithms, how to optimize these components, and how these data-driven components can improve the accuracy and efficiency of traditional algorithms, from both the empirical and theoretical perspectives.Ph.D