t-SNE Grid Search Resulting Diverse Projections
[Sorting Projections According to Metric:
Quality Metrics Average (QMA)
Neighborhood Hit (NH)
Trustworthiness (T)
Continuity (C)
Stress (S)
Shepard Diagram Correlation (SDC)
]
Confirm
Close
Parameters
[Mode:
Grid Search
Single Set
]
Data sets
Diabetes
Breast Cancer
Iris
SPECTF
Gaussian Clusters
Upload File
Factory reset
Perplexity
30
Learning rate
1
Max iterations
500
Load execution
Cache distances
Store execution
Execute new t-SNE analysis
Projections Provenance
[Sorting Projections According to Metric:
Quality Metrics Average (QMA)
Neighborhood Hit (NH)
Trustworthiness (T)
Continuity (C)
Stress (S)
Shepard Diagram Correlation (SDC)
]
Optimize Selection
Interaction Modes
t-SNE Points Exploration
Group Selection
Dimension Correlation
Reset Filters
Overview
Visual Mapping
Density
Color-encoding
Size-encoding
Remaining cost
Size-encoding
Correlation measurement
Distance
KNN
Correlation threshold (%)
K-value (KNN)
50
10
Points radius scaling
3
Hide annotations' controllers
Reveal annotations
Annotation:
Add annotation
Continue the analysis
Delete annotations
Shepard Heatmap
[Visualization:
Shepard Heatmap
Shepard Diagram
]
Dimension Correlation
Min. Visible Correlation: #
0.0
Density and Remaining Cost Distributions
Min. Visible Cost Rate: #
1
Neighborhood Preservation
[Visualization:
Bar Chart
Difference Bar Chart
Line Plot
Difference Line Plot
]
Adaptive Parallel Coordinates Plot
-->