Log10 (Shear modulus - GPa) - log_g_vrh

ARFS Top features

ARFS selected descriptors

---

Correlation analysis

Distance correlation

Dependency graphs

Feature learnability

---

Model performance

5-Fold CV Metrics overview

RF - MATMINER

	train_rmse	test_rmse	train_errors	test_errors	train_r2	test_r2
mean	0.05998	0.16438	0.03464	0.09632	0.98278	0.869
min	0.0521	0.1349	0.0338	0.0917	0.9804	0.7852
max	0.0644	0.2225	0.0364	0.1036	0.9867	0.8986
std	0.00414072	0.0315851	0.000939361	0.00429017	0.00212735	0.043152

RF - MATMINER+LOBSTER

	train_rmse	test_rmse	train_errors	test_errors	train_r2	test_r2
mean	0.05944	0.15536	0.03328	0.09108	0.98304	0.88182
min	0.0491	0.1206	0.0322	0.0847	0.9803	0.7944
max	0.0646	0.2177	0.0352	0.0994	0.9882	0.9267
std	0.00536865	0.0369465	0.00101666	0.00569048	0.00277604	0.0483531

MODNet - MATMINER

	train_rmse	test_rmse	train_errors	test_errors	train_r2	test_r2
mean	0.08042	0.12398	0.02318	0.06704	0.96684	0.9215
min	0.0348	0.0925	0.0202	0.062	0.9577	0.8215
max	0.0947	0.2028	0.0274	0.0718	0.9941	0.9574
std	0.0228767	0.0403218	0.00243343	0.00397522	0.0137634	0.0503536

MODNet - MATMINER+LOBSTER

	train_rmse	test_rmse	train_errors	test_errors	train_r2	test_r2
mean	0.08218	0.1216	0.02634	0.06566	0.96568	0.92412
min	0.0388	0.0897	0.0222	0.0607	0.9557	0.824
max	0.0984	0.2014	0.0338	0.0707	0.9926	0.96
std	0.0219141	0.0410796	0.00424292	0.00427018	0.013602	0.0505217

Corrected resampled t-test on 10-fold CV

Summary

	t_stat	p_value	significance_stars	d_av	rel_improvement	percent_folds_improved
RF	7.56265	1.72926e-05	***	0.583957	4.31678	100
MODNet	1.01519	0.168263		0.211847	2.51596	80

---

Model Explainer

PFI

SHAP

---

SISSO Models

Rung 1

1D descriptor

\[\begin{align*} & log\_g\_vrh = 3.086 \\ & -0.007\left(ElementProperty_{MagpieData\_maximum\_MendeleevNumber} \\ * AverageBondLength_{mean\_Average\_bond\_length}\right) \end{align*}\]

2D descriptor

\[\begin{align*} & log\_g\_vrh = 2.679 \\ & + 1.127\left(MaximumPackingEfficiency_{max\_packing\_efficiency} \\ * ElementProperty_{MagpieData\_mean\_NsValence}\right) \\ & -0.009\left(ElementProperty_{MagpieData\_maximum\_MendeleevNumber} \\ * AverageBondLength_{mean\_Average\_bond\_length}\right) \end{align*}\]

Rung 2

1D descriptor

\[\begin{align*} & log\_g\_vrh = 3.472 \\ & -0.006\left(\frac{ \left(ElementProperty_{MagpieData\_maximum\_MendeleevNumber} * AverageBondLength_{mean\_Average\_bond\_length}\right) }{ \left(\sqrt[3]{ MaximumPackingEfficiency_{max\_packing\_efficiency} }\right) } \right) \end{align*}\]

2D descriptor

\[\begin{align*} & log\_g\_vrh = 2.423 \\ & + 0.29\left(\left(\sqrt[3]{ ElementProperty_{MagpieData\_mean\_NUnfilled} }\right) \\ + \left(\left|AGNIFingerPrint_{std\_dev\_AGNI\_dir\_y\_eta\_2.89e+00} - ElementProperty_{MagpieData\_mean\_NsValence}\right|\right)\right) \\ & -0.006\left(\frac{ \left(ElementProperty_{MagpieData\_maximum\_MendeleevNumber} AverageBondLength_{mean\_Average\_bond\_length}\right) }{ \left(\sqrt[3]{ MaximumPackingEfficiency_{max\_packing\_efficiency} }\right) } \right) \end{align*}\]

---

Misc

ARFS n-iter convergence checks

MAE/ fold from 10-fold CV

Alternative visual summary of input data for t-test