Summary of 3_Linear

<< Go back

Logistic Regression (Linear)

Validation

Optimized metric

logloss

Training time

14.0 seconds

Metric details

C1 C2 C3 C4 C5 nan accuracy macro avg weighted avg logloss
precision 0.825 0.848101 0.666667 0.833333 1 0.833333 0.838028 0.834406 0.83647 0.515071
recall 0.767442 0.87013 0.571429 1 1 1 0.838028 0.868167 0.838028 0.515071
f1-score 0.795181 0.858974 0.615385 0.909091 1 0.909091 0.838028 0.847954 0.836144 0.515071
support 43 77 7 5 5 5 0.838028 142 142 0.515071

Confusion matrix

Predicted as C1 Predicted as C2 Predicted as C3 Predicted as C4 Predicted as C5 Predicted as nan
Labeled as C1 33 10 0 0 0 0
Labeled as C2 7 67 2 0 0 1
Labeled as C3 0 2 4 1 0 0
Labeled as C4 0 0 0 5 0 0
Labeled as C5 0 0 0 0 5 0
Labeled as nan 0 0 0 0 0 5

Learning curves

Learning curves

Coefficients

Coefficients learner #1

C1 C2 C3 C4 C5 nan
intercept 1.89661 4.73311 0.313168 -1.85347 -2.43957 -2.64985
quartier -0.0690232 -0.520492 0.280769 -0.197011 0.175582 0.330176
site -0.582602 0.359342 0.494725 0.297185 -0.293456 -0.275193
cote_voirie -0.057919 -0.00634577 -0.265585 0.170897 -0.0815646 0.240517
matricule_arbre 0.232195 -0.152501 -0.5127 0.115912 0.167134 0.14996
genre_arbre 0.224655 -0.155951 0.354036 -0.218611 -0.185036 -0.0190935
espece_arbre -0.127532 -0.106983 -0.275406 0.595506 0.330809 -0.416395
situation 0.347032 -0.308958 0.339237 0.22492 -0.31169 -0.290541
type_sol 0.334279 -0.610421 -1.08511 0.532669 -0.163595 0.992174
surf_permeable 0.749115 -0.129183 -0.689116 0.578937 -0.118665 -0.391089
date_plantation 0.770019 -0.242307 -0.438837 0.182992 0.118448 -0.390315
classe_age 0.269413 -0.131004 0.346802 -0.231565 0.0344866 -0.288132
hauteur -0.422751 0.30564 -0.130745 0.0744087 -0.41052 0.583967
classe_hauteur 0.197638 0.333466 -0.0657908 -0.335988 -0.415089 0.285764
diametre -0.00137278 -0.249939 -0.0491795 -0.338085 0.277887 0.360689
circonference (en cm) -0.00137278 -0.249939 -0.0491795 -0.338085 0.277887 0.360689
classe_circonference -0.600672 0.0687286 0.274753 -0.363808 0.246348 0.37465
port_arbre 0.14454 -0.00263849 0.568355 -0.197853 -0.125059 -0.387345
vigueur_pousse 0.778623 -0.25484 0.201818 0.101664 -0.493926 -0.333338
plaie_collet 0.602204 -0.50791 -0.195289 0.0165047 -0.141935 0.226425
rejet_tronc 0.305942 -0.0402351 0.575186 -0.0749786 -0.33891 -0.427004
tuteurage_arbre -0.211183 1.05375 -0.328805 -0.393688 -0.109677 -0.0103937
canisse_arbre 0.872519 -0.590594 -0.876563 0.68941 -0.0736011 -0.0211699
plaie_tronc 0.309558 0.402056 -0.382171 -0.409827 0.0126527 0.0677316
champignon_houppier -0.0317644 -0.329102 -0.0882225 0.511329 0.044008 -0.106249
fissure_houppier 0.00144877 0.282041 0.262076 -0.303087 -0.243152 0.000673708
bois_mort_houppier 0.78076 -0.0663505 -0.449878 -0.424198 -0.00514947 0.164816
plaie_houppier 0.509451 0.388228 -0.352651 -0.455034 -0.185651 0.0956575
esperance_maintien -1.80137 -0.419574 0.886968 1.21003 0.810899 -0.686958
contrainte -0.0987119 -0.16551 -0.0705857 -0.0418059 0.769036 -0.392422
date_diagnostic -0.73924 -0.592679 0.805366 -0.023401 0.372522 0.177432
prescription_1 -0.0438974 0.347947 0.427542 -0.152804 -0.836765 0.257978
prescription_2 0.356812 0.198394 -0.635027 0.0896253 0.0341129 -0.0439177
Long -0.624515 0.89516 -0.374953 0.206672 0.118174 -0.220537
Lat 0.246282 -0.176627 0.681353 -0.326147 -0.0619319 -0.362929

Permutation-based Importance

Permutation-based Importance

SHAP Importance

SHAP Importance

SHAP Dependence plots

Dependence C1 (Fold 1)

SHAP Dependence from fold 1

Dependence C2 (Fold 1)

SHAP Dependence from fold 1

Dependence C3 (Fold 1)

SHAP Dependence from fold 1

Dependence C4 (Fold 1)

SHAP Dependence from fold 1

Dependence C5 (Fold 1)

SHAP Dependence from fold 1

Dependence nan (Fold 1)

SHAP Dependence from fold 1

SHAP Decision plots

Worst decisions for selected sample 1 (Fold 1)

SHAP worst decisions from Fold 1

Worst decisions for selected sample 2 (Fold 1)

SHAP worst decisions from Fold 1

Worst decisions for selected sample 3 (Fold 1)

SHAP worst decisions from Fold 1

Worst decisions for selected sample 4 (Fold 1)

SHAP worst decisions from Fold 1

Best decisions for selected sample 1 (Fold 1)

SHAP best decisions from Fold 1

Best decisions for selected sample 2 (Fold 1)

SHAP best decisions from Fold 1

Best decisions for selected sample 3 (Fold 1)

SHAP best decisions from Fold 1

Best decisions for selected sample 4 (Fold 1)

SHAP best decisions from Fold 1

<< Go back