Answered step by step
Verified Expert Solution
Link Copied!

Question

1 Approved Answer

solve the a , b , c , d , e for me and for c , d , e question python is import numpy

solve the a,b,c,d,e for me and for c,d,e question python is import numpy as np
import matplotlib.pyplot as plt
import random
def f(x):
return 1*x**4-5*x**2-3*x
def data_generator(N=1000, seed=1234):
np.random.seed(seed)
x = np.random.uniform(low=-3., high=3., size=[N,1])
epsilon = np.random.normal(loc=0.0, scale=5.0, size=[N,1])
y = f(x)+ epsilon
return (x, y)
def plot_groundtruth(ax):
x = np.linspace(-3.1,3.1,1000)
ax.plot(x, f(x), linestyle='-', color='black')
def plot_data(ax, x, y):
ax.scatter(x, y, marker='o', color='gray', alpha=0.7)
def transformation(x, K):
phi = np.zeros([len(x), K])
for k in range(K):
phi[:,[k]]= x**k
return phi
def f_predict(w, phi):
return np.matmul(phi, w)
def evaluation(y, y_hat): #
return np.mean(np.sqrt((y-y_hat)**2))
### FILL IN ############################################################
def optimize_CF(phi, y):
result = ### FILL IN
return result
def compute_gradient(w, phi, y):
'''
Note. ignore the effect of sigma (of the noise) as this will be capture in the step-size
'''
grad = ### FILL IN
return 1/len(phi)* grad
# do not use tolerance and make it run until the final iteration
def optimize_GD(phi, y, step_size =1e-4, iteration =20000):
error = np.zeros([iteration])
w = np.zeros([phi.shape[1],1]) #initialization
itr =0
while(itr iteration):
grad = compute_gradient(w, phi, y)
w = ### FILL IN
error[itr]= ###FILL IN using function "evaluation()"
itr +=1
print('w_opt_GD: {}'.format(w))
return w, error
#########################################################################
# do not touch this...
def run_problem3_cde():
K =5
x, y = data_generator(N=500, seed=1234)
fig, ax = plt.subplots()
plot_groundtruth(ax)
plot_data(ax, x,y)
ax.set_ylabel('f(x)', fontsize=15)
ax.set_xlabel('x', fontsize=15)
ax.grid()
ax.set_ylim([-23,53])
phi = transformation(x, K)
w_opt_CF = optimize_CF(phi, y)
iteration =20000
step_size =2e-4
### run GD
w_opt_GD, error_GD = optimize_GD(phi, y, step_size = step_size, iteration = iteration)
z = np.linspace(-3.1,3.1,1000).reshape([-1,1])
### plot f(x; w_opt_CF)
ax.plot(z, f_predict(w_opt_CF, transformation(z, K)), linestyle='-', color='red')
### plot f(x; w_opt_GD)
ax.plot(z, f_predict(w_opt_GD, transformation(z, K)), linestyle='-', color='blue')
plt.legend(['Ground Truth', 'Data Points', 'Closed Form', 'GD'], loc=1)
plt.show()
plt.close()
plt.figure()
plt.plot(np.arange(iteration), error_GD, alpha =0.7, color='blue')
plt.grid()
plt.xlabel('iteration', fontsize=15)
plt.ylabel('MSE', fontsize=15)
plt.show()
plt.close()
### do not touch this...
def run_problem3_g(K):
x, y = data_generator(N=50, seed=1234)
fig, ax = plt.subplots()
plot_groundtruth(ax)
plot_data(ax, x,y)
ax.set_ylabel('f(x)', fontsize=15)
ax.set_xlabel('x', fontsize=15)
ax.grid()
ax.set_ylim([-23,53])
phi = transformation(x, K)
w_opt_CF = optimize_CF(phi, y)
z = np.linspace(-3.1,3.1,1000).reshape([-1,1])
### plot f(x; w_opt_CF)
ax.plot(z, f_predict(w_opt_CF, transformation(z, K)), linestyle='-', color='red')
plt.legend(['Ground Truth', 'Data Points', 'K={}'.format(K)], loc=1)
plt.show()
plt.close()
if __name__=="__main__":
run_problem3_cde()
run_problem3_g(K=20) #you should change K to compare polynomial regression w
image text in transcribed

Step by Step Solution

There are 3 Steps involved in it

Step: 1

blur-text-image

Get Instant Access to Expert-Tailored Solutions

See step-by-step solutions with expert insights and AI powered tools for academic success

Step: 2

blur-text-image_2

Step: 3

blur-text-image_3

Ace Your Homework with AI

Get the answers you need in no time with our AI-driven, step-by-step assistance

Get Started

Recommended Textbook for

Database Security XI Status And Prospects

Authors: T.Y. Lin, Shelly Qian

1st Edition

0412820900, 978-0412820908

More Books

Students also viewed these Databases questions

Question

i need 1 8 7 .

Answered: 1 week ago

Question

What is meant by organisational theory ?

Answered: 1 week ago

Question

What is meant by decentralisation of authority ?

Answered: 1 week ago

Question

Briefly explain the qualities of an able supervisor

Answered: 1 week ago

Question

Define policy making?

Answered: 1 week ago

Question

Define co-ordination?

Answered: 1 week ago