dataset = load_analytic_data("dataset.csv")
# Encoding the labels
genres = dataset.iloc[:, -1] # Last column
encoder = LabelEncoder()
labels = encoder.fit_transform(genres)
# Scaling the features
scaler = StandardScaler() # MinMaxScaler() can be also used
features = scaler.fit_transform(np.array(dataset.iloc[:, :-1], dtype=float))
# Dividing dataset into training and testing sets
# 80to20 split
x_train, x_test, y_train, y_test = train_test_split(features,
labels,
test_size=0.2)
# Create knn model
model = KNeighborsClassifier(n_neighbors=9, weights="distance")
# Training
model.fit(x_train, y_train)
# Testing
accuracy = model.score(x_test, y_test)
print(accuracy)
# Save model
save_sklearn_model(model, "knn.sk")
dataset = load_analytic_data("dataset.csv")
# Encoding the labels
genres = dataset.iloc[:, -1] # Last column
encoder = LabelEncoder()
labels = encoder.fit_transform(genres)
# Scaling the features
scaler = StandardScaler() # MinMaxScaler() can be also used
features = scaler.fit_transform(np.array(dataset.iloc[:, :-1], dtype=float))
# Dividing dataset into training and testing sets
# 80to20 split
x_train, x_test, y_train, y_test = train_test_split(features,
labels,
test_size=0.2)
# Create knn model
model = SVC() # NuSVC, SVR, NuSVR, LinearSVC, LinearSVR and OneClassSVM
# Training
model.fit(x_train, y_train)
# Testing
accuracy = model.score(x_test, y_test)
print(accuracy)
# Save model
save_sklearn_model(model, "svm.sk")
# Initialize array
x3 = np.empty((0, n))
# Append x3
for x1, x2 in x:
temp = sin(x1) * x2
x3 = np.append(x3, temp)
# Build the X matrix
X = np.insert(x, 2, x3, axis=1)
# Fit the model
linear_regressor.fit(X, y)
# Get the parameters
theta0 = linear_regressor.intercept_
theta1, theta2, theta3 = linear_regressor.coef_
print(
"\nThe parameter values are: theta0 = {}, theta1 = {}, theta2 = {}, theta3 {}."
.format(theta0, theta1, theta2, theta3))
# Make the predictions of the model
y_pred = linear_regressor.predict(X)
# Print the prediction
MSE = evaluate_predictions(y_pred, y)
print("Task 1 Linear Rregression Model MSE: {}\n".format(MSE))
# Save the model
save_sklearn_model(linear_regressor, '../deliverable/Linear_Regression.pickle')
from sklearn.linear_model import LinearRegression
from sklearn.preprocessing import FunctionTransformer
from sklearn.model_selection import train_test_split
from sklearn.pipeline import make_pipeline
from utils import save_sklearn_model
from deliverable.run_model import load_data
from deliverable.run_model import transform
if __name__ == "__main__":
X, y = load_data("../data/data.npz")
X_train, _, y_train, _ = train_test_split(X, y, train_size=0.85, test_size=0.15, random_state=1)
model = make_pipeline(FunctionTransformer(transform), LinearRegression())
model.fit(X_train, y_train)
save_sklearn_model(model, "../deliverable/t1.pickle")
from sklearn.linear_model import LinearRegression
if __name__ == '__main__':
"""
Store arrays from data.npz in x and y.
Split data into train and test set.
Distinguish between features and labels.
Note: since ordinary least squares is invariant, there is no need for standardization.
"""
x, y = utils.load_data("../data/data.npz")
X = np.column_stack((x, np.sin(x[:, 0]) * x[:, 1]))
train_set, test_set, train_labels, test_labels = train_test_split(X, y, test_size=0.2, random_state=42)
"""
Linear model.
From sklearn train the linear regression model on the train set.
"""
regr = LinearRegression()
regr.fit(train_set, train_labels)
"""
Save the model in deliverable.
"""
utils.save_sklearn_model(regr, '../deliverable/linear_regression.pickle')