Design And Realization Of Data Visualization Theory On Spatio-Temporal Data

In the era of big data,data visualization is one of the fastest growing and most eye-catching fields in the data science.Looking back at the historical development process,it is a derivative technology and composite field that depends on social development,technological progress and aesthetic improvement,and has extremely high application value and practical significance.This thesis takes statistics as the fundamental foothold and extends to computer science,psychology,aesthetics,cartography,history and other disciplines,realizing the integration of theory and technology.It starts from the theoretical research of data visualization,including visual perception,judgment indicators and principles,etc.,and reviews the history of visualization development and classic cases from a statistical perspective to demonstrate the important role of data visualization on data-driven decisions.This thesis outlines the related concepts of data science and the key role of visualization,emphasizes four major breakthrough issues faced by statistics and the immeasurable potential of data visualization.It summarizes the commonly used clustering and anomaly detection algorithms and presents them visually.Then,it focuses on geographic,time-varying and spatio-temporal information visualization.From the perspective of the creators,visualization is the translation of data and information,stressing "faithfulness and elegance",so the theoretical system and evaluation criteria provide important constraints and guidance.From the perspective of the observers,visualization is a portrait of data,which needs to be effective,accurate,while data mining and other technologies can maximize the extraction of data information.Therefore,this thesis can provide new ideas and references for data analysts and other data-related practitioners.