\u09b6\u09a4 \u09ad\u09be\u0997 \u09b8\u09a0\u09bf\u0995 \u09a8\u09be\u0989 \u09b9\u09a4\u09c7 \u09aa\u09be\u09b0\u09c7\u0964 <\/p>\n\n\n\n
<\/p>\n\n\n\n
Theory reference: https:\/\/www.cmpe.boun.edu.tr\/~ethem\/i2ml\/slides\/v1-1\/i2ml-chap5-v1-1.pdf . This is an approximate solution to start with. Understand the theory and then adjust\/fix\/improve<\/p>\n\n\n\n
import numpy as np<\/p>\n\n\n\n
import random
<\/p>\n\n\n\n
print(‘Iterations: rows: Please enter the number of samples for each variable\/dimension’)<\/p>\n\n\n\n
n_number_of_samples_rows = int(input())
<\/p>\n\n\n\n
print(‘Columns: Dimensions: Please enter number of variables’)<\/p>\n\n\n\n
m_number_of_variables_cols = int(input())
<\/p>\n\n\n\n
# initialize the input data variables<\/p>\n\n\n\n
print(m_number_of_variables_cols, n_number_of_samples_rows)<\/p>\n\n\n\n
X = np.zeros((n_number_of_samples_rows, m_number_of_variables_cols))
<\/p>\n\n\n\n
#Y_actually_R = np.zeros((n_number_of_samples_rows, m_number_of_variables_cols))<\/p>\n\n\n\n
Y_actually_R = np.zeros((n_number_of_samples_rows * 1)) #m_number_of_variables_cols
<\/p>\n\n\n\n
#generate random data<\/p>\n\n\n\n
for n in range (n_number_of_samples_rows):<\/p>\n\n\n\n
for m in range(m_number_of_variables_cols):<\/p>\n\n\n\n
X[n,m] = random.random()<\/p>\n\n\n\n
Y_actually_R[n] = random.random()
<\/p>\n\n\n\n
print(“X”) <\/p>\n\n\n\n
print(X)
<\/p>\n\n\n\n
print(“Y in row format”)<\/p>\n\n\n\n
print(Y_actually_R)
<\/p>\n\n\n\n
print(“Y in column\/vector format”)<\/p>\n\n\n\n
Y_actually_R = np.transpose(Y_actually_R) #transpose<\/p>\n\n\n\n
print(Y_actually_R)<\/p>\n\n\n\n
<\/p>\n\n\n\n
#convert to matrix<\/p>\n\n\n\n
X = np.matrix(X)<\/p>\n\n\n\n
Y_actually_R = np.matrix(Y_actually_R)
<\/p>\n\n\n\n
print(‘————-matrix-print—X and then Y’)<\/p>\n\n\n\n
print(X)
<\/p>\n\n\n\n
print(‘Y matrix’)<\/p>\n\n\n\n
print(Y_actually_R)<\/p>\n\n\n\n
<\/p>\n\n\n\n
#THE EQUATION: steps to calculate W matrix: w = (((X.Transpose) * X).invert) * X.Transpose * r
<\/p>\n\n\n\n
#transpose X<\/p>\n\n\n\n
X_transpose = np.transpose(X)<\/p>\n\n\n\n
print(‘X_transpose’)<\/p>\n\n\n\n
print(X_transpose)<\/p>\n\n\n\n
<\/p>\n\n\n\n
#(X.Transpose) * X) of [w = (((X.Transpose) * X).invert) * X.Transpose * r]<\/p>\n\n\n\n
w_parameters = np.dot(X_transpose, X) #does np.multiply work? probably need shapoing\/reshaping<\/p>\n\n\n\n
print(‘first dot’)<\/p>\n\n\n\n
print(w_parameters)<\/p>\n\n\n\n
<\/p>\n\n\n\n
#(X.Transpose) * X).invert<\/p>\n\n\n\n
w_parameters = np.linalg.inv(w_parameters)<\/p>\n\n\n\n
print(‘inverted’)<\/p>\n\n\n\n
print(w_parameters)
<\/p>\n\n\n\n
#(((X.Transpose) * X).invert) * X.Transpose <\/p>\n\n\n\n
w_parameters = np.dot(w_parameters, X_transpose)<\/p>\n\n\n\n
print(‘2nd dot’)<\/p>\n\n\n\n
print(w_parameters)<\/p>\n\n\n\n
<\/p>\n\n\n\n
#(((X.Transpose) * X).invert) * X.Transpose * r<\/p>\n\n\n\n
#Y_actually_R = np.transpose(Y_actually_R)<\/p>\n\n\n\n
w_parameters = np.dot(w_parameters, np.transpose(Y_actually_R)) #np.dot( np.transpose(w_parameters), np.transpose(Y_actually_R))<\/p>\n\n\n\n
#two times transpose – redundant. actually, we could avoid both transpose of Y upto this point <\/p>\n\n\n\n
<\/p>\n\n\n\n
print(‘w_matrix’)<\/p>\n\n\n\n
print(w_parameters)<\/p>\n\n\n\n
w_matrix = w_parameters
<\/p>\n\n\n\n
#sum of ( rt – w0 – w1x1 – w2x2 ….. wd * xd ) rt = Y_Actually_R[t 1….N][variable_1….d]<\/p>\n\n\n\n
#d eqv to m_number_of_variables_cols — i used m for that<\/p>\n\n\n\n
#E(w 0 ,w 1 ,…,w d |X )
<\/p>\n\n\n\n
#Should it be a matrix or just one total sum? I assume one total sum<\/p>\n\n\n\n
error_matrix = np.zeros(m_number_of_variables_cols)<\/p>\n\n\n\n
error_sum = 0
<\/p>\n\n\n\n
#calculate error<\/p>\n\n\n\n
Y_actually_R = np.transpose(Y_actually_R) #np.array(Y_actually_R)<\/p>\n\n\n\n
for n in range(n_number_of_samples_rows):<\/p>\n\n\n\n
sum = 0<\/p>\n\n\n\n
for m in range(m_number_of_variables_cols):<\/p>\n\n\n\n
sum = Y_actually_R[m] <\/p>\n\n\n\n
#2nd part ie w1x1 w2x2 wdxd of the equation: sum of ( rt – w0 – w1x1 – w2x2 ….. wd * xd )<\/p>\n\n\n\n
wpart = w_matrix[0] <\/p>\n\n\n\n
for ii in range(1,m_number_of_variables_cols): # d = m_number_of_variables_cols, sum of w1x1, w2x2 to wdxd <\/p>\n\n\n\n
wpart += w_matrix[ii] * X[n,m] #+ w_matrix[2][m] * X[n][m]<\/p>\n\n\n\n
#sum = sum – wpart<\/p>\n\n\n\n
sum = pow( (sum – wpart), 2)<\/p>\n\n\n\n
error_matrix[m] = 0.5 * pow ( (sum – wpart), 2) <\/p>\n\n\n\n
error_sum += sum #error_matrix[m] #pow ( (sum – wpart), 2)<\/p>\n\n\n\n
error_sum = 0.5 * error_sum <\/p>\n\n\n\n
print(‘error matrix if supposed to be = number of variables’) <\/p>\n\n\n\n
print(error_matrix)
<\/p>\n\n\n\n
print(‘Error if supposed to be one number i.e. sum of all errors’)<\/p>\n\n\n\n
print(error_sum)
<\/p>\n","protected":false},"excerpt":{"rendered":"
\u09ae\u09c7\u09b6\u09bf\u09a8 \u09b2\u09be\u09b0\u09a8\u09bf\u0982 \u0983 Implement: Multivariate Regression: Python \u09b6\u09a4 \u09ad\u09be\u0997 \u09b8\u09a0\u09bf\u0995 \u09a8\u09be\u0989 \u09b9\u09a4\u09c7 \u09aa\u09be\u09b0\u09c7\u0964 Theory reference: https:\/\/www.cmpe.boun.edu.tr\/~ethem\/i2ml\/slides\/v1-1\/i2ml-chap5-v1-1.pdf . This is an approximate solution to start with. Understand the theory and then adjust\/fix\/improve import numpy as np import random print(‘Iterations: rows: Please enter the number of samples for each variable\/dimension’) n_number_of_samples_rows = int(input()) print(‘Columns: Dimensions: Please enter … <\/p>\n