Machine_learning_pure_python/02_BinaryLogReg.py at master · sophronesis/Machine_learning_pure_python · GitHub

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
from ml_algo import BinaryLogisticRegression, circle_cluster
import numpy as np
from matplotlib import pylab as plt
from random import seed
seed(12345)

def main():

	#init random cluster of dots
	num_train = 400
	num_test = 200
	red_dots = np.array([circle_cluster(10,10,10) for i in range(num_train//2)])
	blue_dots = np.array([circle_cluster(20,0,10) for i in range(num_train//2)])
	Xtrain = np.concatenate((red_dots,blue_dots))
	ytrain = np.array([0 for _ in range(num_train//2)]+[1 for _ in range(num_train//2)])

	red_dots_test = np.array([circle_cluster(10,10,10) for i in range(num_test//2)])
	blue_dots_test = np.array([circle_cluster(20,0,10) for i in range(num_test//2)])
	Xtest = np.concatenate((red_dots_test,blue_dots_test))
	ytest = np.array([0 for _ in range(num_test//2)]+[1 for _ in range(num_test//2)])

	#use logistic regression, calculating on test data and output metrics
	model = BinaryLogisticRegression(0.03,2)
	model.fit(Xtrain,ytrain,100,1e-5)
	model2 = BinaryLogisticRegression(0.03,2,momentum=0)
	model2.fit(Xtrain,ytrain,100,1e-5)
	threshold = lambda x: 0 if x<0.5 else 1
	confmat = np.zeros((2,2))
	confusion_dots = [[[],[]],[[],[]]]
	for i in range(len(Xtest)):
		result = threshold(model.predict(Xtest[i]))
		confmat[result][ytest[i]]+=1
		confusion_dots[result][ytest[i]].append(Xtest[i])
	print('Confusion matrix:')
	print(confmat)
	accuracy = (confmat[0][0]+confmat[1][1])/np.sum(confmat)
	precision = confmat[0][0]/(confmat[0][0]+confmat[1][0])
	recall = confmat[0][0]/(confmat[0][0]+confmat[0][1])
	f1score = 2*precision*recall/(precision+recall)
	print('\nAccuracy:\t{}\nPrecision:\t{}\nRecall:  \t{}\nF1 score:\t{}'.format(accuracy,precision,recall,f1score))

	#plotting graph
	f, plots = plt.subplots(2, 1)
	plots[0].plot(*np.array(confusion_dots[0][0]).T,'r.')
	plots[0].plot(*np.array(confusion_dots[1][0]).T,'rx')
	plots[0].plot(*np.array(confusion_dots[0][1]).T,'bx')
	plots[0].plot(*np.array(confusion_dots[1][1]).T,'b.')
	plots[0].set_title('Classification task')
	plots[1].plot(model.cost_func_log)
	plots[1].plot(model2.cost_func_log)
	plots[1].set_title('Cost function (blue curve - with momentum grad)')
	plt.show()

if __name__ == '__main__':
	main()