Table 3.

UMAP parameters

Parameter	Abbreviation	Levels
Learning rate	LR	8
Metric for high dimensional space distances calculation	Metr	8
Number of nearest neighbors assumed at local level	LC	5
Dispersion of points on manifold	MiD	5
Size of neighboring sample points in manifold estimation	NN	6
During optimization, ratio of negative samples per positive example	NSR	3
Negative samples penalization while optimizing in low dimension	RS	4
Ratio of fuzzy set operations to obtain global fuzzy simplicial sets	Mix	5
Spread out scale of embedded points	Sp	5