7 Models

# Logistic Regression
from sklearn.linear_model import LogisticRegression

params_lr = {'penalty': 'l1', 'solver':'liblinear'}

model_lr = LogisticRegression(**params_lr)
model_lr, accuracy_lr, roc_auc_lr, coh_kap_lr, tt_lr = run_model(model_lr, X_train, y_train, X_test, y_test)

# Decision Tree
from sklearn.tree import DecisionTreeClassifier

params_dt = {'max_depth': 16,
             'max_features': "sqrt"}

model_dt = DecisionTreeClassifier(**params_dt)
model_dt, accuracy_dt, roc_auc_dt, coh_kap_dt, tt_dt = run_model(model_dt, X_train, y_train, X_test, y_test)

# Neural Network
from sklearn.neural_network import MLPClassifier

params_nn = {'hidden_layer_sizes': (30,30,30),
             'activation': 'logistic',
             'solver': 'lbfgs',
             'max_iter': 500}

model_nn = MLPClassifier(**params_nn)
model_nn, accuracy_nn, roc_auc_nn, coh_kap_nn, tt_nn = run_model(model_nn, X_train, y_train, X_test, y_test)

# Random Forest
from sklearn.ensemble import RandomForestClassifier

params_rf = {'max_depth': 16,
             'min_samples_leaf': 1,
             'min_samples_split': 2,
             'n_estimators': 100,
             'random_state': 12345}

model_rf = RandomForestClassifier(**params_rf)
model_rf, accuracy_rf, roc_auc_rf, coh_kap_rf, tt_rf = run_model(model_rf, X_train, y_train, X_test, y_test)

# Light GBM
import lightgbm as lgb
params_lgb ={'colsample_bytree': 0.95, 
         'max_depth': 16, 
         'min_split_gain': 0.1, 
         'n_estimators': 200, 
         'num_leaves': 50, 
         'reg_alpha': 1.2, 
         'reg_lambda': 1.2, 
         'subsample': 0.95, 
         'subsample_freq': 20}

model_lgb = lgb.LGBMClassifier(**params_lgb)
model_lgb, accuracy_lgb, roc_auc_lgb, coh_kap_lgb, tt_lgb = run_model(model_lgb, X_train, y_train, X_test, y_test)

# Catboost
!pip install catboost
import catboost as cb
params_cb ={'iterations': 50,
            'max_depth': 16}

model_cb = cb.CatBoostClassifier(**params_cb)
model_cb, accuracy_cb, roc_auc_cb, coh_kap_cb, tt_cb = run_model(model_cb, X_train, y_train, X_test, y_test, verbose=False)

# XGBoost
import xgboost as xgb
params_xgb ={'n_estimators': 500,
            'max_depth': 16}

model_xgb = xgb.XGBClassifier(**params_xgb)
model_xgb, accuracy_xgb, roc_auc_xgb, coh_kap_xgb, tt_xgb = run_model(model_xgb, X_train, y_train, X_test, y_test)