Table 1 Identification of the important features for predicting EFA values.
From: Discovery of high-entropy ceramics via machine learning
Predictor rank | Model | |
|---|---|---|
Stoichiometric attributes | CALPHAD | |
1 | avg(ionic character) | avg(ionic character) |
2 | min(electrons) | Liquidus temperature* |
3 | avg. dev(s-valence electrons) | range(electronegativity) |
4 | max(atomic weight) | avg. dev(d-valence electrons) |
5 | max(covalent radius) | max(atomic weight) |
6 | fwm(covalent radius) | fwm(f-valence electrons) |
7 | range(Mendeleev number) | max(covalent radius) |
8 | avg. dev(melting temp) | max(unfilled valence electrons) |
9 | fwm(unfilled s-valence) | fwm(covalent radius) |
10 | fwm(f-electrons) | range(unfilled valence electrons) |
