kaggle入门第一天：机器学习基础 感知器与PLA

#coding:utf-8
from select import kevent

import numpy as np
import matplotlib.pyplot as plt

def draw(trainingData, w, round, x):
plt.figure('Round'+str(round))
drawLine(w)
drawTrainingData(trainingData)

if x is not None:
plt.scatter(x[1], x[2], s= 400, c = 'red', marker=r'$\bigodot$')
plt.xticks([])
plt.yticks([])
plt.show()

def drawLine( w):
w = w.transpose().tolist()[0]
x1 = [x1 for x1 in xrange(80)]
x2 = [ (w[0] + w[1] * i)/(-w[2]) for i in x1]
plt.plot(x1, x2)

def drawTrainingData(trainingData):
pointSize = 100
positive_color = 'red'
positive_marker = 'o'
negative_clor = 'blue'
negative_marker = 'x'

positive_x1 = []
positive_x2 = []
negative_x1 = []
negative_x2 = []
for x, y in trainingData:
x = x.transpose().tolist()[0]

if y == 1:
positive_x1.append(x[1])
positive_x2.append(x[2])
elif y == -1:
negative_x1.append(x[1])
negative_x2.append(x[2])

plt.scatter(positive_x1, positive_x2, s= pointSize, c = positive_color, marker = positive_marker)
plt.scatter(negative_x1, negative_x2, s= pointSize, c = negative_clor, marker=negative_marker)

def vector(l):
return np.mat(l).transpose()

def trainingDataPreprocess(trainingData):
'''Add x0 dimension & transform to np.mat object'''
processedTrainingData = [ (vector([1, sample[0], sample[1]]), sample[2])  for sample in trainingData]
return processedTrainingData

def PLA(trainingData):
w = np.mat([1,2127,205]).transpose() # Step 1: 向量w赋初值

k = 0 # 第k轮计数
while True:
k += 1

(status, x, y) = noMistakePoint(trainingData, w)
draw(trainingData, w, k, x) # 画图
if status == 'YES': # Step 2: 切分正确，学习完成
return w
else:
w = w + y*x # Step 3: 修正w

sign = lambda x:1 if x > 0 else -1 if x < 0 else -1
def mSign(m):
'''判断某个矩阵的[0][0]元素正负.大于0返回1，否则返回-1'''
x = m.tolist()[0][0]
return 1 if x > 0 else -1 if x < 0 else -1

def noMistakePoint(training_data, w):
'''训练数据中是否有点被切分错误'''
status = 'YES'
for (x, y) in training_data:
if mSign(w.transpose() * x) <> sign(y):
status = 'NO'
return (status, x, y)

return status, None, None

if __name__=="__main__":

trainingData = [
[10, 300, -1],
[15, 377, -1],
[50, 137, 1],
[65, 92 , 1],
[45, 528, -1],
[61, 542, 1],
[26, 394, -1],
[37, 703, -1],
[39, 244, 1],
[41, 398, 1],
[53, 495, 1],
[32, 119, 1],
[24, 577, -1],
[56, 412, 1]
]

processedTrainingData = trainingDataPreprocess(trainingData)
w = PLA(processedTrainingData)