Self-organizing Map (SOM) Based Data Navigation for Identifying Shape Similarities of Graphic Logos

Wang, Florence Y.; Takatsuka, Masahiro

doi:10.1007/s11063-014-9375-4

Cited by 2 publications

(1 citation statement)

References 9 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…A notable exception is the CIELAB color model [6], which was designed to be largely perceptually uniform, meaning that the Euclidean distance between any two points should correspond to their relative perceptual difference. This makes it particularly useful for our application, since we want to visualize positions as colors, similar to what has previously been tried for self-organizing maps [7]. CIELAB represents colors using three values: perceptual lightness (L*) and two pairs of complementary colors, namely green–red (a*) and blue–yellow (b*).…”

Section: Mainmentioning

confidence: 99%

U-CIE [/juː ‘siː/]: Color encoding of high-dimensional data

Koutrouli

Morris

Jensen

2021

Preprint

View full text Add to dashboard Cite

U-CIE is a visualization method that encodes arbitrary high-dimensional data as colors using a combination of dimensionality reduction and the CIELAB color space to retain the original structure to the extent possible. We illustrate its broad applicability by visualizing single-cell data on a protein network and metagenomic data on a world map and on scatter plots. U-CIE is available as a web resource at https://u-cie.jensenlab.org/ and as an R package.

show abstract

Section: Mainmentioning

confidence: 99%

U-CIE [/juː ‘siː/]: Color encoding of high-dimensional data

Koutrouli

Morris

Jensen

2021

Preprint

View full text Add to dashboard Cite

show abstract

Identification of Influential Nodes for Drone Swarm Based on Graph Neural Networks

Wang

Zhuang

Xie

2021

Neural Process Lett

View full text Add to dashboard Cite

Accurately tracking the influential nodes in swarm is the key to improve the control or defense against the drone combat. But the identification of influential nodes is intractable due to hardly capturing the pattern information of the running drones based on the traditional Graph Theory. To tackle this issue, in this paper, we propose a novel framework to identify the influential nodes, which includes Latent Interaction Graph Extraction (LIGE), Degree Centrality based Influential Inference (DCII) and Formation Stability Coefficient based Re-ranking (FSCR). LIGE is employed to reconstruct the latent interaction graph from the dynamic trajectory data by exploiting the encoder in Neural Relational Inference. DCII invokes the traditional graph model and measures the significance of nodes. In the re-ranking module FSCR, we introduce a novel metric, the formation stability coefficient, to measure swarm formation stability and re-ranking the candidates. The experimental results on the Drone swarm datasets demonstrate that our framework can effectively identify the influential nodes of drone swarm. The extracting time in LIGE is relatively long, which makes it difficult to recognize the core node within the limit time. By optimizing the encoder structure, the extraction time can be reduced to at least twice the original.

show abstract

Self-organizing Map (SOM) Based Data Navigation for Identifying Shape Similarities of Graphic Logos

Cited by 2 publications

References 9 publications

U-CIE [/juː ‘siː/]: Color encoding of high-dimensional data

U-CIE [/juː ‘siː/]: Color encoding of high-dimensional data

Identification of Influential Nodes for Drone Swarm Based on Graph Neural Networks

Contact Info

Product

Resources

About