Table 5. Algorithm Comparison Summary: ULI vs. SDM.
Pros/Cons Summary | |
---|---|
Sparse Diffusion Maps | Unweighted Landmark Isomap |
Con: Only accurately reconstructs cubes | Pro: Can accurately reconstruct non-cubes |
Con: Biased towards exterior points | Pro: Not biased towards exterior points |
Pro: Is robust to problems that have high similarity between some far-away points. | Con: Is generally not robust to problems that have high similarity between some far-away points. |