t-SNE: 10D to 2D Projection

Example 4: High-Dimensional Clusters (10D -> 2D)

80 points in 4 distinct 10-dimensional clusters. t-SNE should clearly separate the clusters in 2D.

Feature Selection for Distance Calculation:

Available Features:

Selected Features:

Rescaling: (Changes which variables determine similarity)

Color nodes by:

Perplexity: 10.0 (5-100, typically 15-50)

Learning Rate: 200.0 (10-1000, typically 100-500)

Convergence threshold:

Max iterations:

Early exaggeration iters:

Exaggeration factor:

Iteration: 0
         
Movement (last iteration): 0.000
Status: Ready

Aspect Ratio: 0.6

Zoom: 1.0x

Tip: Drag nodes to manually reposition. "Run to Convergence" uses early exaggeration for the configured number of iterations; "Step (small)" never uses exaggeration; "Exaggerated Step" always uses it.

Data: highdim_data.csv

t-SNE Examples This page was created using JSPlots.jl v0.6.3.