使用tensorflow实现线性svm

Ava ·

更新时间:2024-11-14

· 502 次阅读

本文实例为大家分享了tensorflow实现线性svm的具体代码，供大家参考，具体内容如下

简单方法：


import tensorflow as tf
import numpy as np
from matplotlib import pyplot as plt
def placeholder_input():
  x=tf.placeholder('float',shape=[None,2],name='x_batch')
  y=tf.placeholder('float',shape=[None,1],name='y_batch')
  return x,y
def get_base(_nx, _ny):
  _xf = np.linspace(x_min, x_max, _nx)
  _yf = np.linspace(y_min, y_max, _ny)
  xf1, yf1 = np.meshgrid(_xf, _yf)
  n_xf,n_yf=np.hstack((xf1)),np.hstack((yf1))
  return _xf, _yf,np.c_[n_xf.ravel(), n_yf.ravel()]
x_data=np.load('x.npy')
y1=np.load('y.npy')
y_data=np.reshape(y1,[200,1])
step=10000
tol=1e-3
x,y=placeholder_input()
w = tf.Variable(np.ones([2,1]), dtype=tf.float32, name="w_v")
b = tf.Variable(0., dtype=tf.float32, name="b_v")
y_pred =tf.matmul(x,w)+b 
y_predict =tf.sign( tf.matmul(x,w)+b )
# cost = ∑_(i=1)^N max⁡(1-y_i⋅(w⋅x_i+b),0)+1/2 + 0.5 * ‖w‖^2
cost = tf.nn.l2_loss(w)+tf.reduce_sum(tf.maximum(1-y*y_pred,0))
train_step = tf.train.AdamOptimizer(0.01).minimize(cost)
with tf.Session() as sess:
  sess.run(tf.global_variables_initializer())
  for i in range(step):
    sess.run(train_step,feed_dict={x:x_data,y:y_data})
    y_p,y_p1,loss,w_value,b_value=sess.run([y_predict,y_pred,cost,w,b],feed_dict={x:x_data,y:y_data})
x_min, y_min = np.minimum.reduce(x_data,axis=0) -2
x_max, y_max = np.maximum.reduce(x_data,axis=0) +2
xf, yf , matrix_= get_base(200, 200)
#xy_xf, xy_yf = np.meshgrid(xf, yf, sparse=True)
z=np.sign(np.matmul(matrix_,w_value)+b_value).reshape((200,200))
plt.pcolormesh(xf, yf, z, cmap=plt.cm.Paired)
for i in range(200):
  if y_p[i,0]==1.0:
    plt.scatter(x_data[i,0],x_data[i,1],color='r')
  else:
    plt.scatter(x_data[i,0],x_data[i,1],color='g')
plt.axis([x_min,x_max,y_min ,y_max])
#plt.contour(xf, yf, z)
plt.show()

进阶：


import tensorflow as tf
import numpy as np
from matplotlib import pyplot as plt
class SVM():
  def __init__(self):
    self.x=tf.placeholder('float',shape=[None,2],name='x_batch')
    self.y=tf.placeholder('float',shape=[None,1],name='y_batch')
    self.sess=tf.Session()
  @staticmethod
  def get_base(self,_nx, _ny):
    _xf = np.linspace(self.x_min, self.x_max, _nx)
    _yf = np.linspace(self.y_min, self.y_max, _ny)
    n_xf, n_yf = np.meshgrid(_xf, _yf)
    return _xf, _yf,np.c_[n_xf.ravel(), n_yf.ravel()]
  def readdata(self):
    x_data=np.load('x.npy')
    y1=np.load('y.npy')
    y_data=np.reshape(y1,[200,1])
    return x_data ,y_data
  def train(self,step,x_data,y_data):
    w = tf.Variable(np.ones([2,1]), dtype=tf.float32, name="w_v")
    b = tf.Variable(0., dtype=tf.float32, name="b_v")
    self.y_pred =tf.matmul(self.x,w)+b 
    cost = tf.nn.l2_loss(w)+tf.reduce_sum(tf.maximum(1-self.y*self.y_pred,0))
    train_step = tf.train.AdamOptimizer(0.01).minimize(cost)
    self.y_predict =tf.sign( tf.matmul(self.x,w)+b )
    self.sess.run(tf.global_variables_initializer())
    for i in range(step):      
      self.sess.run(train_step,feed_dict={self.x:x_data,self.y:y_data})
      self.y_predict_value,self.w_value,self.b_value,cost_value=self.sess.run([self.y_predict,w,b,cost],feed_dict={self.x:x_data,self.y:y_data})
      print('**********cost=%f***********'%cost_value)
  def predict(self,y_data):    
    correct = tf.equal(self.y_predict_value, y_data)
    precision=tf.reduce_mean(tf.cast(correct, tf.float32)) 
    precision_value=self.sess.run(precision)
    return precision_value
  def drawresult(self,x_data):
    self.x_min, self.y_min = np.minimum.reduce(x_data,axis=0) -2
    self.x_max, self.y_max = np.maximum.reduce(x_data,axis=0) +2
    xf, yf , matrix_= self.get_base(self,200, 200)
    w_value=self.w_value
    b_value=self.b_value
    print(w_value,b_value)
    z=np.sign(np.matmul(matrix_,self.w_value)+self.b_value).reshape((200,200))
    plt.pcolormesh(xf, yf, z, cmap=plt.cm.Paired)
    for i in range(200):
      if self.y_predict_value[i,0]==1.0:
        plt.scatter(x_data[i,0],x_data[i,1],color='r')
      else:
        plt.scatter(x_data[i,0],x_data[i,1],color='g')
    plt.axis([self.x_min,self.x_max,self.y_min ,self.y_max])
    #plt.contour(xf, yf, z)
    plt.show()     
svm=SVM()
x_data,y_data=svm.readdata()
svm.train(5000,x_data,y_data)
precision_value=svm.predict(y_data)
svm.drawresult(x_data)

没有数据的可以用这个


import tensorflow as tf
import numpy as np
from matplotlib import pyplot as plt
class SVM():
  def __init__(self):
    self.x=tf.placeholder('float',shape=[None,2],name='x_batch')
    self.y=tf.placeholder('float',shape=[None,1],name='y_batch')
    self.sess=tf.Session()
  def creat_dataset(self,size, n_dim=2, center=0, dis=2, scale=1, one_hot=False):
    center1 = (np.random.random(n_dim) + center - 0.5) * scale + dis
    center2 = (np.random.random(n_dim) + center - 0.5) * scale - dis
    cluster1 = (np.random.randn(size, n_dim) + center1) * scale
    cluster2 = (np.random.randn(size, n_dim) + center2) * scale
    x_data = np.vstack((cluster1, cluster2)).astype(np.float32)
    y_data = np.array([1] * size + [-1] * size)
    indices = np.random.permutation(size * 2)
    x_data, y_data = x_data[indices], y_data[indices]
    y_data=np.reshape(y_data,(y_data.shape[0],1))
    if not one_hot:
      return x_data, y_data
    y_data = np.array([[0, 1] if label == 1 else [1, 0] for label in y_data], dtype=np.int8)
    return x_data, y_data
  @staticmethod
  def get_base(self,_nx, _ny):
    _xf = np.linspace(self.x_min, self.x_max, _nx)
    _yf = np.linspace(self.y_min, self.y_max, _ny)
    n_xf, n_yf = np.meshgrid(_xf, _yf)
    return _xf, _yf,np.c_[n_xf.ravel(), n_yf.ravel()]
#  def readdata(self):
#    
#    x_data=np.load('x.npy')
#    y1=np.load('y.npy')
#    y_data=np.reshape(y1,[200,1])
#    return x_data ,y_data
  def train(self,step,x_data,y_data):
    w = tf.Variable(np.ones([2,1]), dtype=tf.float32, name="w_v")
    b = tf.Variable(0., dtype=tf.float32, name="b_v")
    self.y_pred =tf.matmul(self.x,w)+b 
    cost = tf.nn.l2_loss(w)+tf.reduce_sum(tf.maximum(1-self.y*self.y_pred,0))
    train_step = tf.train.AdamOptimizer(0.01).minimize(cost)
    self.y_predict =tf.sign( tf.matmul(self.x,w)+b )
    self.sess.run(tf.global_variables_initializer())
    for i in range(step):
      index=np.random.permutation(y_data.shape[0])
      x_data1, y_data1 = x_data[index], y_data[index]
      self.sess.run(train_step,feed_dict={self.x:x_data1[0:50],self.y:y_data1[0:50]})
      self.y_predict_value,self.w_value,self.b_value,cost_value=self.sess.run([self.y_predict,w,b,cost],feed_dict={self.x:x_data,self.y:y_data})
      if i%1000==0:print('**********cost=%f***********'%cost_value)
  def predict(self,y_data):    
    correct = tf.equal(self.y_predict_value, y_data)
    precision=tf.reduce_mean(tf.cast(correct, tf.float32)) 
    precision_value=self.sess.run(precision)
    return precision_value, self.y_predict_value
  def drawresult(self,x_data):
    self.x_min, self.y_min = np.minimum.reduce(x_data,axis=0) -2
    self.x_max, self.y_max = np.maximum.reduce(x_data,axis=0) +2
    xf, yf , matrix_= self.get_base(self,200, 200)
    print(self.w_value,self.b_value)
    z=np.sign(np.matmul(matrix_,self.w_value)+self.b_value).reshape((200,200))
    plt.pcolormesh(xf, yf, z, cmap=plt.cm.Paired)
    for i in range(x_data.shape[0]):
      if self.y_predict_value[i,0]==1.0:
        plt.scatter(x_data[i,0],x_data[i,1],color='r')
      else:
        plt.scatter(x_data[i,0],x_data[i,1],color='g')
    plt.axis([self.x_min,self.x_max,self.y_min ,self.y_max])
#    plt.contour(xf, yf, z)
    plt.show()     
svm=SVM()
x_data,y_data=svm.creat_dataset(size=200, n_dim=2, center=0, dis=4, one_hot=False)
svm.train(5000,x_data,y_data)
precision_value,y_predict_value=svm.predict(y_data)
svm.drawresult(x_data)

您可能感兴趣的文章:win10下python3.5.2和tensorflow安装环境搭建教程win10下tensorflow和matplotlib安装教程python3.6.3安装图文教程 TensorFlow安装配置方法tensorflow实现简单逻辑回归Tensorflow使用支持向量机拟合线性回归TensorFlow实现iris数据集线性回归TensorFlow实现模型评估使用TensorFlow实现SVMTensorFlow Session使用的两种方法小结C++调用tensorflow教程

svm tensorflow

1024 个赞