Machine Learning Python:

**How do I get rid of lines of code bolded in class Ridge Regression and replace them with functions sgd(stochastic gradient descent) and predict? I want to use the functions sgd and predict for the Ridge Regression class.

Code:

class RidgeRegression() :

def __init__( self, learning_rate, iterations, l2_penality ) :

self.learning_rate = learning_rate

self.iterations = iterations

self.l2_penality = l2_penality

# Function for model training

def fit( self, X, Y ) :

# no_of_training_examples, no_of_features

self.m, self.n = X.shape

# weight initialization

self.W = np.zeros( self.n )

self.b = 0

self.X = X

self.Y = Y

# gradient descent learning

for i in range( self.iterations ) :

self.update_weights()

return self

# Helper function to update weights in gradient descent

def update_weights( self ) :

Y_pred = self.predict( self.X )

# calculate gradients

dW = ( - ( 2 * ( self.X.T ).dot( self.Y - Y_pred ) ) +

( 2 * self.l2_penality * self.W ) ) / self.m

db = - 2 * np.sum( self.Y - Y_pred ) / self.m

# update weights

self.W = self.W - self.learning_rate * dW

self.b = self.b - self.learning_rate * db

return self

# Hypothetical function h( x )

def predict( self, X ) :

return X.dot( self.W ) + self.b

def sgd(train_data,learning_rate,n_iter,k,divideby): # Initially we will keep our W and B as 0 as per the Training Data w=np.zeros(shape=(1,train_data.shape[1]-1)) b=0 cur_iter=1 while(cur_iter<=n_iter):

# We will create a small training data set of size K temp=train_data.sample(k) # We create our X and Y from the above temp dataset y=np.array(temp['Electrical Energy Output']) x=np.array(temp.drop('Electrical Energy Output',axis=1)) # We keep our initial gradients as 0 w_gradient=np.zeros(shape=(1,train_data.shape[1]-1)) b_gradient=0 for i in range(k): # Calculating gradients for point in our K sized dataset prediction=np.dot(w,x[i])+b w_gradient=w_gradient+(-2)*x[i]*(y[i]-(prediction)) b_gradient=b_gradient+(-2)*(y[i]-(prediction)) #Updating the weights(W) and Bias(b) with the above calculated Gradients w=w-learning_rate*(w_gradient/k) b=b-learning_rate*(b_gradient/k) # Incrementing the iteration value cur_iter=cur_iter+1 #Dividing the learning rate by the specified value learning_rate=learning_rate/divideby return w,b #Returning the weights and Bias

def predict(x,w,b): y_pred=[] for i in range(len(x)): y=np.asscalar(np.dot(w,x[i])+b) y_pred.append(y) return np.array(y_pred)