Homec4science
Diffusion Optical_Trapping_ML (master) Browse

R13271/GRAM_measurements/optuna_objectives.py
995d2644e47cmaster

Last Change
Raw File

optuna_objectives.py
View Options

# optuna_objectives.py
import optuna
from xgboost import XGBClassifier
import numpy as np
#objective function for optuna optimization for both gram and bacteria classification
class ObjectiveXGB:
def __init__(self, X_train, X_test, y_train, y_test, X_train_bacteria, X_test_bacteria, y_train_bacteria, y_test_bacteria):
self.X_train = X_train
self.X_test = X_test
self.y_train = y_train
self.y_test = y_test
self.X_train_bacteria = X_train_bacteria
self.X_test_bacteria = X_test_bacteria
self.y_train_bacteria = y_train_bacteria
self.y_test_bacteria = y_test_bacteria
def objective_xgb_gram(self, trial):
param = {
'n_estimators': trial.suggest_int('n_estimators', 100, 1000),
'max_depth': trial.suggest_int('max_depth', 3, 10),
'learning_rate': trial.suggest_float('learning_rate', 0.01, 0.3, log=True),
'subsample': trial.suggest_float('subsample', 0.5, 1.0),
'colsample_bytree': trial.suggest_float('colsample_bytree', 0.5, 1.0),
'gamma': trial.suggest_float('gamma', 1e-8, 10.0, log=True),
'reg_alpha': trial.suggest_float('reg_alpha', 1e-8, 10.0, log=True),
'reg_lambda': trial.suggest_float('reg_lambda', 1e-8, 10.0, log=True),
'min_child_weight': trial.suggest_int('min_child_weight', 1, 10),
}
xgb_gram = XGBClassifier(**param, use_label_encoder=False, eval_metric='logloss')
xgb_gram.fit(self.X_train, self.y_train, eval_set=[(self.X_test, self.y_test)], early_stopping_rounds=10, verbose=False)
preds = xgb_gram.predict(self.X_test)
accuracy = np.mean(preds == self.y_test)
return accuracy
def objective_xgb_bacteria(self, trial):
param = {
'objective': 'multi:softmax',
'num_class': len(np.unique(self.y_train_bacteria)),
'n_estimators': trial.suggest_int('n_estimators', 100, 1000),
'max_depth': trial.suggest_int('max_depth', 3, 10),
'learning_rate': trial.suggest_float('learning_rate', 0.01, 0.3, log=True),
'subsample': trial.suggest_float('subsample', 0.5, 1.0),
'colsample_bytree': trial.suggest_float('colsample_bytree', 0.5, 1.0),
'gamma': trial.suggest_float('gamma', 1e-8, 10.0, log=True),
'reg_alpha': trial.suggest_float('reg_alpha', 1e-8, 10.0, log=True),
'reg_lambda': trial.suggest_float('reg_lambda', 1e-8, 10.0, log=True),
'min_child_weight': trial.suggest_int('min_child_weight', 1, 10),
}
xgb_bacteria = XGBClassifier(**param, use_label_encoder=False, eval_metric='mlogloss')
xgb_bacteria.fit(self.X_train_bacteria, self.y_train_bacteria, eval_set=[(self.X_test_bacteria, self.y_test_bacteria)], early_stopping_rounds=10, verbose=False)
preds = xgb_bacteria.predict(self.X_test_bacteria)
accuracy = np.mean(preds == self.y_test_bacteria)
return accuracy
c4science · Help