Created
March 8, 2020 05:02
-
-
Save amogh112/ba93061a53d4511fcb136adaa0a91a5a to your computer and use it in GitHub Desktop.
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
import numpy as np | |
def generateData(a, b, num_samples): | |
data_x = np.random.uniform(a, b, num_samples) | |
data_y = np.random.uniform(a, b, num_samples) | |
data = np.array(list(zip(data_x, data_y))) | |
labels = np.array([-1 if datap[0] <= datap[1] else 1 for datap in data]) | |
return data, labels | |
def evaluate(datap, weights): | |
result = datap @ weights.T | |
if result > 0: | |
return 1 | |
else: | |
return -1 | |
def trainPerceptron(max_iter, data, labels,hinge=False): | |
weights = np.zeros(2) | |
bias = np.zeros(1) | |
for i in range(max_iter): | |
for i, x in enumerate(data): | |
if hinge: | |
act = (x @ weights.T) + bias[0] | |
else: | |
act = (x @ weights.T) | |
y = labels[i] | |
if y * act <= 0: | |
weights = weights + y * x | |
bias = bias + y | |
return weights, bias | |
def predictPerceptron(weights, bias, x, hinge): | |
if hinge: | |
act = x @ weights.T + bias | |
else: | |
act = x @ weights.T | |
result = np.ones_like(act) | |
result[np.where(act <= 0)] = -1 | |
return result | |
def mainPerceptron(hinge): | |
# hinge = False | |
x, y = generateData(0, 1, 20) | |
weights, bias = trainPerceptron(1000, x, y, hinge) | |
import matplotlib.pyplot as plt | |
for i,dp in enumerate(x): | |
if y[i] == 1: | |
plt.scatter(*dp, color='orange') | |
else: | |
plt.scatter(*dp, color='blue') | |
# plt.show() | |
test_x, test_y = generateData(0, 1, 1000) | |
result = predictPerceptron(weights,bias,test_x,hinge) | |
correct = np.sum(result == test_y) | |
accuracy = correct/test_y.shape[0] | |
print(accuracy) | |
if __name__ == "__main__": | |
for i in range(20): | |
mainPerceptron(hinge=False) |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment