基于机器学习的增强型可视化理论与方法

本项目将研究基于机器学习的增强型可视化理论与方法。项目的核心目标是利用机器学习方法从群体智能和专家知识中学习的智能，从可视表达中学习和转换有意义的特征表达，从而增强交互可视化的效率和质量，克服常规可视化方法过于依赖用户的操作、强于形象而弱于抽象等挑战。同时，通过对机器学习的特征表达转换为可视表达，提升机器学习方法的可解释性和易用性。项目将以电力系统仿真数据的分析为实例，验证理论与方法的有效性。

This project proposes to study the foundations and approaches of enhanced visualization based on machine learning. The key idea is to learn and transform meaningful feature representations to/from visual representations by leveraging well-established machine learning approaches from crowd intelligence and expert knowledge. This scheme could enhance the efficiency and quality of interactive visualization, and address the problems of the heavy dependency of the user performance as well as the strengths on visual cognition and the weakness on logic abstraction of current visualization theory and approaches. In the meantime, the interpretability and usability of machine learning approaches could be empowered by transforming feature representations to visual representations. We will demonstrate the effectiveness of the proposed scheme by a suite of examples for analyzing the simulation data of power systems.