MAD uses a multi-output Gaussian process with co-regionalization kernel to jointly map crystal structures via NMF and optimize electrical resistance in a closed-loop, identifying promising PCMs in the Mn-Sb-Te ternary within 25 iterations for a reported seven-fold speedup.
Title resolution pending
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cond-mat.mtrl-sci 1years
2026 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
Real-time Multi-instrument Autonomous Discovery of Novel Phase-change Memory Materials
MAD uses a multi-output Gaussian process with co-regionalization kernel to jointly map crystal structures via NMF and optimize electrical resistance in a closed-loop, identifying promising PCMs in the Mn-Sb-Te ternary within 25 iterations for a reported seven-fold speedup.