def mlp_regressor():
csv_path = os.path.join(os.path.dirname(os.path.abspath(__file__)), '..',
'data.csv')
dataframe = pandas.read_csv(csv_path, header=0)
X = dataframe.drop('ИН', axis=1)
y = dataframe['ИН']
X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.25)
model = MLPRegressor(max_iter=200, solver='lbfgs')
scores = cross_validate(model,
X_train,
y_train,
cv=10,
return_train_score=True)
predict_start = time()
predictions = cross_val_predict(model, X_test, y_test, cv=10)
predict_end = time()
predict_time = predict_end - predict_start
model.fit(X_train, y_train)
# print(model.coefs_)
print(model.n_layers_)
print(model.n_outputs_)
method_results(model, scores, y_test, predictions,
[np.shape(a) for a in model.coefs_], predict_time, X_train)
def linear_regression():
csv_path = os.path.join(os.path.dirname(os.path.abspath(__file__)), '..', 'data.csv')
dataframe = pandas.read_csv(csv_path, header=0)
X = dataframe.drop('ИН', axis=1)
y = dataframe['ИН']
X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.25)
model = LinearRegression()
scores = cross_validate(model, X_train, y_train, cv=10, return_train_score=True)
predict_start = time()
predictions = cross_val_predict(model, X_test, y_test, cv=10)
predict_end = time()
predict_time = predict_end - predict_start
model.fit(X_train, y_train)
method_results(model, scores, y_test, predictions, model.coef_.shape[0], predict_time, X_train)
def random_forest():
csv_path = os.path.join(os.path.dirname(os.path.abspath(__file__)), '..', 'data.csv')
dataframe = pandas.read_csv(csv_path, header=0)
X = dataframe.drop('ИН', axis=1)
y = dataframe['ИН']
X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.25)
model = RandomForestRegressor(n_estimators=20)
scores = cross_validate(model, X_train, y_train, cv=10, return_train_score=True)
predict_start = time()
predictions = cross_val_predict(model, X_test, y_test, cv=10)
predict_end = time()
predict_time = predict_end - predict_start
model.fit(X_train, y_train)
method_results(model, scores, y_test, predictions, model.feature_importances_.shape[0], predict_time, X_train)
def decision_tree():
csv_path = os.path.join(os.path.dirname(os.path.abspath(__file__)), '..', 'data.csv')
dataframe = pandas.read_csv(csv_path, header=0)
X = dataframe.drop('ИН', axis=1)
y = dataframe['ИН']
X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.25)
model = tree.DecisionTreeRegressor(max_depth=100)
scores = cross_validate(model, X_train, y_train, cv=10, return_train_score=True)
predict_start = time()
predictions = cross_val_predict(model, X_test, y_test, cv=10)
predict_end = time()
predict_time = predict_end - predict_start
model.fit(X_train, y_train)
method_results(model, scores, y_test, predictions, model.feature_importances_.shape[0], predict_time, X_train)
graph_data = tree.export_graphviz(model, out_file=None, filled=True)
graph = graphviz.Source(graph_data)
graph.render("children", view=True)
def k_nearest():
csv_path = os.path.join(os.path.dirname(os.path.abspath(__file__)), '..',
'data.csv')
dataframe = pandas.read_csv(csv_path, header=0)
X = dataframe.drop('ИН', axis=1)
y = dataframe['ИН']
X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.25)
model = neighbors.KNeighborsRegressor(n_neighbors=10)
scores = cross_validate(model,
X_train,
y_train,
cv=10,
return_train_score=True)
predict_start = time()
predictions = cross_val_predict(model, X_test, y_test, cv=10)
predict_end = time()
predict_time = predict_end - predict_start
method_results(model, scores, y_test, predictions, None, predict_time,
X_train)
def svm():
csv_path = os.path.join(os.path.dirname(os.path.abspath(__file__)), '..',
'data.csv')
dataframe = pandas.read_csv(csv_path, header=0)
X = dataframe.drop('ИН', axis=1)
y = dataframe['ИН']
rbX = RobustScaler()
X = rbX.fit_transform(X)
rbY = RobustScaler()
y = rbY.fit_transform(y.values.reshape(-1, 1))
X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.25)
# kernel: rbf, sigmoid, linear, poly gamma=0.1, C=100, epsilon=0.1
model = SVR(kernel='rbf', gamma='auto')
scores = cross_validate(model,
X_train,
y_train.ravel(),
cv=10,
return_train_score=True)
predict_start = time()
predictions = cross_val_predict(model,
rbX.transform(X_test),
y_test.ravel(),
cv=10)
predict_end = time()
predict_time = predict_end - predict_start
model.fit(X_train, y_train.ravel())
params_shape = model.support_vectors_.shape[
0] * model.support_vectors_.shape[1]
method_results(model, scores, y_test, predictions, params_shape,
predict_time, X_train)