Modified empirical formulas and machine learning for α-decay systematics

Abstract Latest experimental and evaluated α -decay half-lives between 82 ⩽ Z 118 have been used to modify two empirical formulas: (i) Horoi scaling law (2004 J. Phys. G: Nucl. Part. 30 945), Sobiczewski formula (2005 Acta Pol. B 36 3095) by adding asymmetry dependent terms ( I 2 ) refitting of the coefficients. The results these modified formulas are found with significant improvement while compared other 21 formulas, and, therefore, predict more precision in unknown superheavy region. spontaneous fission (SF) half-life proposed Bao et al (2015 42 085101) is further using ground-state shell-plus-pairing correction taken from FRDM-2012 latest SF 118. Using contest probed for nuclei within range 112 consequently probable decay modes estimated. Potential chains 286−302 Og 287−303 119 (168 N 184: island stability) analyzed which be excellent agreement available data. In addition, four different machine learning models: XGBoost, random forest, decision trees, multilayer perceptron (MLP) neural network train a predictor prediction. prediction XGBoost MLP along our predictions obtained above-mentioned formulas.