Supercompress Outperforms Low-Discrepancy Point Methods in Machine Learning Data Compression Tests
Experimental benchmarks reveal that adaptive response clustering markedly outperforms Quasi-Monte Carlo point sets for compressing machine learning datasets.

