Differences

This shows you the differences between two versions of the page.

Link to this comparison view

Next revision		Previous revision
highlights:2025:colorpca [2025/09/08 08:28] – created Sohaib Ghanihighlights:2025:colorpca [2025/09/08 08:55] (current) – Sohaib Ghani
Line 1: Line 1:
-====== Evolution of Extreme Dust Events in 3D Environment ======+====== ColorPCA: Scalable Colored Dimensionality Reduction for Unlabeled High-Dimensional Data ======
  
 <WRAP group> <WRAP group>
 <WRAP third column> <WRAP third column>
  
-{{:wiki:highlights:2025:dustVis.png?500|}}+{{:wiki:highlights:2025:colorpca.png?500|}}
  
 <WRAP center round box 100%> {{:favicon.ico?nolink&0x30}} <fs:26px>** KVL Staff on Project**</fs> <WRAP center round box 100%> {{:favicon.ico?nolink&0x30}} <fs:26px>** KVL Staff on Project**</fs>
  
-{{:icon:twbs:link:w-25:h-auto:person.svg?nolink&}} James Kress \\ +{{:icon:twbs:link:w-25:h-auto:person.svg?nolink&}} Sohaib Ghani \\ 
-{{:icon:twbs:link:w-25:h-auto:envelope-at.svg?nolink&}}  james.kress@kaust.edu.sa \\ +{{:icon:twbs:link:w-25:h-auto:envelope-at.svg?nolink&}}  sohaib.ghani@kaust.edu.sa \\ 
-{{:icon:twbs:link:w-25:h-auto:building.svg?nolink&}} Building 1, Level 0, Office 0120 \\+{{:icon:twbs:link:w-25:h-auto:building.svg?nolink&}} Building 1, Level 0, Office 0121 \\
 </WRAP> </WRAP>
  
Line 25: Line 25:
  
 ===== Overview ===== ===== Overview =====
-KVL in collaboration with Prof. Ibrahim Hoteit's group visualized a simulation of an extreme dust event that happened in Saudi Arabia in July of 2018. With the goal of understanding the formation, progression, and driving factors behind the storm.+KVL in collaboration with Prof. Ibrahim Hoteit's group, designed an interactive technique for dimensionality reduction of high-dimensional data.
  
-Dust is one of the main components of atmospheric particles in desert regions. The concentration, composition, and spatial +Mapping labeled high-dimensional data to colors based on class labels in low-dimensional 
-distribution of these dust particles in the atmosphere vary over time and could significantly impact the weather, climatological +projections is effective for enhancing pattern recognition and reducing misinterpretation of clusters. 
-conditions, radiative forcing and transfer, and ecosystem dynamics. Scientists and decision-makers are interested in analyzing +However, automatic coloring of unlabeled high-dimensional data remains challenging for revealing unknown 
-the evolution of dust events (including their formation, dynamics, and interactions with the environment), understanding the +patterns or class structures in the data. To address this, we propose ColorPCA, a scalable method that 
-main contributing factors and atmospheric conditions that intensify these events and lead to extreme dust events, examining +improves existing dimensionality reduction-based automatic coloring by integrating Principal Component 
-the role of topographic features, and gaining insights into their relationship with global teleconnections+Analysis with alpha compositing. Rather than mapping reduced dimensions to color coordinates, ColorPCA 
 +directly encodes data into color space to enhance pattern discovery in unlabeled datasets. We implemented 
 +ColorPCA in a web-based visual analytics system for interactive exploration and evaluated it through three 
 +case studies using benchmark, simulated, and real-world climate datasets. Additionally, we conducted three 
 +user studies, two with generic users and one with climate domain experts. Comparisons with two state-of- 
 +the-art coloring methods based on PCA and t-SNE demonstrate that ColorPCA improves visual separability 
 +and facilitates deeper insight extraction in high-dimensional data visualization. 
 +  
 +The paper entitled "ColorPCA: Scalable Colored Dimensionality Reduction for Unlabeled High-Dimensional Data" can be accessed [[https://ieeexplore.ieee.org/abstract/document/11021415|here]].
  
-With this work we developed a system to analyze the output from the WRF simulation, Weather Research and Forecasting, simulation run at a fine temporal resolution on Shaheen II. This resulted in a large amount of data that necessitated the design of the visualization environment to quickly process and visualize the simulation results. This system also aids in understanding the interactions between different 
-atmospheric parameters, the impact of terrain surface characteristics, and more, providing a holistic view of the dust events. The paper entitled "Evolution of Extreme Dust Events in 3D Environment" can be accessed [[https://diglib.eg.org/items/d7d0b300-615a-4767-9c1d-fe14cd734147|here]]. 
  
- +{{ wiki:highlights:2025:colorpca-video.mp4?512 |ColorPCA based Dimensionality Reduction}}
-{{ wiki:highlights:2025:final.mp4?512 |Saudi Dust Event WRF Simulation}}+
  
 </WRAP> /* end column */ </WRAP> /* end column */