fix: move super().__init__() before self.regularization in LinearRegression

[Satyam-madeit]

Problem

In the __init__ method of LinearRegression, super().__init__() was being called
after self.regularization was set. The parent class __init__ was overwriting
self.regularization back to None, causing LinearRegression.fit(X, y) to raise:

ValueError: Self.regularization must be assigned

Solution

Moved super().__init__() to the top of __init__, before setting
self.regularization. This ensures the parent class initializes first and
doesn't overwrite the child class attributes.

Before

def __init__(self, n_iterations=100, learning_rate=0.001, gradient_descent=True):
    self.gradient_descent = gradient_descent
    self.regularization = lambda x: 0
    self.regularization.grad = lambda x: 0
    super(LinearRegression, self).__init__(n_iterations=n_iterations,
                                          learning_rate=learning_rate)

After

def __init__(self, n_iterations=100, learning_rate=0.001, gradient_descent=True):
    super(LinearRegression, self).__init__(n_iterations=n_iterations,
                                          learning_rate=learning_rate)
    self.gradient_descent = gradient_descent
    self.regularization = lambda x: 0
    self.regularization.grad = lambda x: 0

Closes #123