Updated model yaml with hyperparameter optimization results

Author: SjB-ngi

scripts/config/model/catboost.yaml

@@ -1,9 +1,9 @@
 name: catboost
 params:
-  iterations: 1000
-  learning_rate: 0.1
-  depth: 6
-  l2_leaf_reg: 3
+  depth: 3
+  iterations: 120
+  l2_leaf_reg: 13.33755638154388
+  learning_rate: 0.055128733444030256
   model_size_reg: 0.5
   rsm: 1
   loss_function: Logloss

scripts/config/model/extra_trees.yaml

@@ -1,7 +1,8 @@
 name: extra_trees
 params:
-  n_estimators: 100
-  max_depth: null
-  min_samples_split: 2
-  min_samples_leaf: 1
+  bootstrap: false
+  max_depth: 12
   max_features: sqrt
+  min_samples_leaf: 2
+  min_samples_split: 6
+  n_estimators: 172

scripts/config/model/hist_gradient_boosting.yaml

@@ -1,17 +1,9 @@
 name: hist_gradient_boosting
 params:
-  learning_rate: 0.1
-  max_iter: 100
-  max_leaf_nodes: 31
-  max_depth: null
-  min_samples_leaf: 20
-  l2_regularization: 0.0
-  max_bins: 255
-  categorical_features: null
-  monotonic_cst: null
-  warm_start: false
-  early_stopping: auto
-  scoring: loss
-  validation_fraction: 0.1
+  l2_regularization: 0.9202038272323202
+  learning_rate: 0.027864170616690123
+  max_iter: 190
+  max_leaf_nodes: 28
+  min_samples_leaf: 42
   n_iter_no_change: 10
   tol: 1e-07

scripts/config/model/knn.yaml

@@ -1,8 +1,5 @@
 name: knn
 params:
-  n_neighbors: 5
-  weights: uniform
-  algorithm: auto
-  leaf_size: 30
+  n_neighbors: 9
   p: 2
-  metric: minkowski
+  weights: uniform

scripts/config/model/lightgbm.yaml

@@ -1,14 +1,14 @@
 name: lightgbm
 params:
-  boosting_type: gbdt
-  objective: binary
-  metric: binary_logloss
-  num_leaves: 31
-  learning_rate: 0.05
-  feature_fraction: 0.9
-  bagging_fraction: 0.8
-  bagging_freq: 5
-  verbose: 0
+  bagging_fraction: 0.8677126844571269
+  bagging_freq: 1
+  feature_fraction: 0.7937635704745494
+  learning_rate: 0.02002821098423169
+  min_child_samples: 43
+  n_estimators: 265
+  num_leaves: 45
+  reg_alpha: 0.17943130140820815
+  reg_lambda: 0.10154626671341846
   min_data_in_leaf: 6
   min_sum_hessian_in_leaf: 11
   n_estimators: 100

scripts/config/model/logistic_regression.yaml

@@ -1,6 +1,5 @@
 name: logistic_regression
 params:
-  penalty: l2
-  C: 1.0
-  solver: lbfgs
-  max_iter: 100
+  C: 0.2895667317295911
+  penalty: l1
+  solver: liblinear

scripts/config/model/random_forest.yaml

@@ -1,7 +1,7 @@
 name: random_forest
 params:
-  n_estimators: 100
-  max_depth: null
-  min_samples_split: 2
-  min_samples_leaf: 1
-  max_features: auto
+  max_depth: 11
+  max_features: log2
+  min_samples_leaf: 4
+  min_samples_split: 5
+  n_estimators: 109

scripts/config/model/svm.yaml

@@ -1,16 +1,5 @@
 name: svm
 params:
-  C: 1.0
-  kernel: rbf
-  degree: 3
+  C: 0.9175201343352734
   gamma: scale
-  coef0: 0.0
-  shrinking: true
-  probability: false
-  tol: 0.001
-  cache_size: 200
-  class_weight: null
-  verbose: false
-  max_iter: -1
-  decision_function_shape: ovr
-  break_ties: false
+  kernel: rbf

scripts/config/model/xgboost.yaml

@@ -1,11 +1,11 @@
 name: xgboost
 params:
-  n_estimators: 179
+  colsample_bytree: 0.7351421288846683
+  learning_rate: 0.07926004251101824
   max_depth: 8
-  learning_rate: 0.1336674424474196
-  subsample: 0.7282263389921348
-  colsample_bytree: 0.8896012201308952
-  reg_alpha: 0.07919227245102388
-  reg_lambda: 0.36469616999655735
+  n_estimators: 96
+  reg_alpha: 0.06779661113890172
+  reg_lambda: 9.088250624476464
+  subsample: 0.9437099479618614
   random_state: 42
   n_jobs: -1