IPAC (Met) GA Models vs Modern Catalogues [Newton, Mann, Gaidos].

Introduction

After the modelling process from GA features, the outcome will be compared against catallogues from Gaidos, Mann and Newton.

Loading predictions from GA based models

Loading recent models

1 final-Gaidos2014-*.tsv contiene muchas columnas. En particular, la primera contiene el nombre de la estrella, la #34 contiene Fe/H y la #35 la temperatura efectiva.

2 final-Mann2015-*.tsv: la primera contiene el nombre de la estrella, la #11 contiene Fe/H y la #8 la temperatura efectiva.

3 final-Newton2014*.tsv: la primera contiene el nombre de la estrella y la #15 contiene Fe/H. En el fichero de IPAC hay dos estrellas que están repetidas: es porque tienen dos estimaciones.

Comparison Fe/H between sources

Gaidos

drmse dmae drmdse dmade
RF_Inf 0.33 0.28 0.25 0.25
GB_Inf 0.34 0.29 0.25 0.25
SVR_Inf 4.50 4.05 4.25 4.25
NNR_Inf 2.25 2.11 2.06 2.06
KNN_Inf 0.80 0.60 0.47 0.47
MARS_Inf 1.48 1.22 1.13 1.13
PLS_Inf 1.33 1.22 1.24 1.24
Rule-Regression_Inf 1.48 1.40 1.47 1.47
RF_10 0.53 0.44 0.38 0.38
GB_10 0.58 0.46 0.41 0.41
SVR_10 0.40 0.31 0.25 0.25
NNR_10 0.56 0.49 0.49 0.49
KNN_10 0.38 0.29 0.23 0.23
MARS_10 0.60 0.50 0.45 0.45
PLS_10 0.71 0.65 0.63 0.63
Rule-Regression_10 0.43 0.33 0.27 0.27
RF_50 0.91 0.83 0.79 0.79
GB_50 0.99 0.91 0.99 0.99
SVR_50 0.64 0.51 0.42 0.42
NNR_50 0.77 0.55 0.46 0.46
KNN_50 1.05 0.92 0.86 0.86
MARS_50 1.06 0.85 0.70 0.70
PLS_50 0.72 0.61 0.55 0.55
Rule-Regression_50 0.55 0.44 0.37 0.37
Chi2_inf 0.63 0.52 0.47 0.47
Chi2_10 0.73 0.57 0.47 0.47
Chi2_50 0.69 0.53 0.45 0.45
ICA_inf 1.17 1.06 1.00 1.00
ICA_10 0.77 0.61 0.52 0.52
ICA_50 1.17 0.94 0.76 0.76
M_teo 0.14 0.11 0.09 0.08
n_muestras
1 75
n_muestras_M_teo
1 26

Mann

drmse dmae drmdse dmade
RF_Inf 0.30 0.25 0.22 0.22
GB_Inf 0.34 0.27 0.25 0.25
SVR_Inf 4.61 4.13 4.16 4.16
NNR_Inf 2.19 2.05 1.98 1.98
KNN_Inf 0.77 0.58 0.39 0.39
MARS_Inf 1.42 1.20 1.14 1.14
PLS_Inf 1.30 1.19 1.15 1.15
Rule-Regression_Inf 1.51 1.43 1.43 1.43
RF_10 0.49 0.42 0.37 0.37
GB_10 0.54 0.44 0.40 0.40
SVR_10 0.40 0.30 0.21 0.21
NNR_10 0.49 0.43 0.42 0.42
KNN_10 0.38 0.30 0.26 0.26
MARS_10 0.55 0.46 0.42 0.42
PLS_10 0.66 0.60 0.61 0.61
Rule-Regression_10 0.42 0.32 0.24 0.24
RF_50 0.90 0.81 0.79 0.79
GB_50 0.96 0.88 0.84 0.84
SVR_50 0.58 0.47 0.50 0.50
NNR_50 0.70 0.52 0.45 0.45
KNN_50 0.97 0.86 0.85 0.85
MARS_50 1.00 0.82 0.74 0.74
PLS_50 0.63 0.52 0.47 0.47
Rule-Regression_50 0.50 0.39 0.35 0.35
Chi2_inf 0.59 0.48 0.40 0.40
Chi2_10 0.77 0.62 0.62 0.62
Chi2_50 0.64 0.52 0.41 0.41
ICA_inf 1.14 0.99 0.87 0.87
ICA_10 0.76 0.56 0.37 0.37
ICA_50 1.20 0.93 0.73 0.73
M_teo 0.08 0.06 0.04 0.04
n_muestras
1 67
n_muestras_M_teo
1 35

Newton

drmse dmae drmdse dmade
RF_Inf 0.30 0.26 0.26 0.26
GB_Inf 0.35 0.29 0.28 0.28
SVR_Inf 3.05 2.55 2.18 2.18
NNR_Inf 2.08 1.93 2.05 2.05
KNN_Inf 0.50 0.41 0.36 0.36
MARS_Inf 1.00 0.71 0.56 0.56
PLS_Inf 1.27 1.18 1.31 1.31
Rule-Regression_Inf 1.25 1.20 1.21 1.21
RF_10 0.67 0.57 0.59 0.59
GB_10 0.80 0.68 0.66 0.66
SVR_10 0.62 0.50 0.44 0.44
NNR_10 0.68 0.61 0.65 0.65
KNN_10 0.51 0.42 0.44 0.44
MARS_10 0.87 0.76 0.74 0.74
PLS_10 0.82 0.75 0.79 0.79
Rule-Regression_10 0.68 0.58 0.59 0.59
RF_50 0.77 0.72 0.75 0.75
GB_50 1.00 0.94 0.99 0.99
SVR_50 0.79 0.70 0.78 0.78
NNR_50 0.80 0.69 0.73 0.73
KNN_50 1.32 1.17 1.29 1.29
MARS_50 0.96 0.87 0.87 0.87
PLS_50 0.80 0.72 0.74 0.74
Rule-Regression_50 0.57 0.49 0.46 0.46
Chi2_inf 0.49 0.41 0.31 0.31
Chi2_10 0.53 0.41 0.28 0.28
Chi2_50 0.53 0.44 0.38 0.38
ICA_inf 0.80 0.73 0.75 0.75
ICA_10 0.47 0.38 0.33 0.33
ICA_50 0.65 0.42 0.20 0.20
M_teo 0.12 0.09 0.08 0.07
n_muestras
1 48
n_muestras_M_teo
1 14