tensorflow2.0——代码实现一元逻辑回归

 

 

 

 

import tensorflow as tf
import numpy as np

x = np.array([1, 2, 3, 4])
y = np.array([0,0,1,1])
w = tf.Variable(1.)
b = tf.Variable(1.)
sigmodX = 1 / (1 + tf.exp(-(w * x + b)))                                    #   sigmod 函数
pre_result = tf.round(sigmodX)                                              #   将结果四舍五入
pre_result2 = tf.where(sigmodX < 0.9,1,0)                                   #   阈值设置
pre_bool = tf.equal(pre_result,y)                                           #   预测值四舍五入后与标记值对比，判断预测是否正确
bool_int = tf.cast(pre_bool,tf.float32)                                     #   将bool转化为0，1
accuary = tf.reduce_mean(bool_int)                                          #   对正确结果数组求平均值就是准确率
loss = -(y * tf.math.log(sigmodX) + (1 - y)* tf.math.log(1 - sigmodX))      #   每个样本的损失值
loss_sum = tf.reduce_sum(loss)                                              #   所有样本的损失总和
loss_mean= tf.reduce_mean(loss)                                             #   所有样本的的平均损失
print('sigmodX:',sigmodX)
print('pre_result:',pre_result)
print('pre_result2:',pre_result2)
print('pre_bool:',pre_bool)
print('bool_int:',bool_int)
print('accuary:',accuary)
print('loss：',loss)
print('loss_sum：',loss_sum)
print('loss_mean：',loss_mean)