Visual Quality of Multidimensional Time Series Projections

Project Overview

This project addresses the challenge of visualizing multidimensional time series data by projecting it into 2D space. We introduce new techniques to manage and display projection errors that may occur during the projection and visualization process. To illustrate the temporal aspect of the data, successive time points are represented as dots connected by lines or curves, highlighting their temporal relationships. However, inaccuracies in projected data points, inconsistent variations in distances between projected time instances, and intersections of connecting lines can result in misinterpretations of the original data. To address these issues, we propose innovative methods to effectively confront these challenges.

My Contribution

I implemented the system and wrote most of the paper. Additionally, I selected, prepared, and analyzed the examples used in our work.

Publications

Results of this project are published in the paper Exploring Visual Quality of Multidimensional Time Series Projections.