class LinearRegressionODE[source]

LinearRegressionODE(lambda_)

Linear Regression with Ordinary Differential Equations

LinearRegressionODE.fit[source]

LinearRegressionODE.fit(X, y)

Args:

X(np.array): (# of samples, # of features)
y(np.array) is an array of shape (n,)

LinearRegressionODE.predict[source]

LinearRegressionODE.predict(X)

Args:

X(np.array): (# of samples, # of features)

Returns:

an array of shape (# of samples,).

lr = LinearRegressionODE(lambda_ = 0.01)

X_train = np.random.rand(100, 3)
y_train = np.random.randint(100)

lr.fit(X = X_train, y = y_train)

lr.beta.shape

(3, 100)

X_test = np.random.rand(10, 3) # (No of sample, No of features) 

y_pred = lr.predict(X_test)

y_pred.shape 

(10, 100)

y_true = np.random.randint(10) 

def MSE(y_pred, y_true):
    """Mean Squred Error"""
    return np.mean((y_pred - y_true) ** 2)

MSE(y_pred, y_true)

5.829916607037568