testing algorithms.py

# Polynomial Linear Regression -->
plr = make_pipeline(PolynomialFeatures(degree=3), linear_model.Ridge())
plr.fit(X_train, y_train)

# Support Vector Regression -->
svr_rbf = SVR(kernel='rbf',C=1e3,gamma=0.1)
grid_param = {'C':[1e0,1e1,1e2,1e3,1e4,1e5],'gamma':np.logspace(-2,2,5)}
grid_svr = GridSearchCV(svr_rbf, cv=KFold(n_splits=10), param_grid=grid_param)
svr = make_pipeline(StandardScaler(),grid_svr)
svr.fit(X_train,y_train)

# Gradient boosting regressor -->
gbr = GradientBoostingRegressor(alpha=0.8,learning_rate=0.06,max_depth=2,min_samples_leaf=2,
                                min_samples_split=2, n_estimators=100, random_state=30)
param_grid_gbr = {'n_estimators':[100,200],'learning_rate':[0.1,0.05,0.02,0.005],'max_depth':[2,4,6],
                   'min_samples_leaf':[3,5,9]}
grid_gbr = GridSearchCV(gbr,param_grid=param_grid_gbr)
gbr = make_pipeline(StandardScaler(),grid_gbr)
gbr.fit(X_train,y_train)

# Descision tree regression -->
dtr = DecisionTreeRegressor(max_depth=5)
param_grid_dtr = {'max_depth':[1,2,3,4,5,6,7]}
grid_dtr = GridSearchCV(dtr, cv=KFold(n_splits=10),param_grid=param_grid_dtr)
dtr = make_pipeline(StandardScaler(),grid_dtr)
dtr.fit(X_train,y_train)