线性回归:通过拟合线性模型的回归系数W =(w_1,…,w_p)来减少数据中观察到的结果和实际结果之间的残差平方和,并通过线性逼近进行预测。

从数学上讲,它解决了下面这个形式的问题:     

LinearRegression()模型在Sklearn.linear_model下,他主要是通过fit(x,y)的方法来训练模型,其中x为数据的属性,y为所属类型。线性模型的回归系数W会保存在他的coef_方法中。

例如:

>>> from sklearn import linear_model
>>> clf = linear_model.LinearRegression()
>>> clf.fit ([[0, 0], [1, 1], [2, 2]], [0, 1, 2])
LinearRegression(copy_X=True, fit_intercept=True, n_jobs=1, normalize=False)
>>> clf.coef_
array([ 0.5, 0.5])

实例:

使用的数据集为Sklearn.dataset.load_diabetes()一个关于糖尿病的数据集。

aaarticlea/png;base64,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" alt="" width="677" height="236" />

为了说明这个回归技术的一个二维图,例子仅仅使用了糖尿病数据集的第一个特征。

代码如下:

# -*- encoding:utf-8 -*-
"""
Line Regression Example
DataBase:diavetes
""" import matplotlib.pyplot as plt
import numpy as np
from sklearn import datasets,linear_model
import time a=time.time()
####加载数据集
diabetes=datasets.load_diabetes() ####仅仅使用一个特征:
diabetes_X=diabetes.data[:,np.newaxis,2] ###s数据划分训练集和测试集
diabetes_X_train=diabetes_X[:-20]
diabetes_X_test=diabetes_X[-20:] ###目标划分为训练集和测试集
diabetes_y_train=diabetes.target[:-20]
diabetes_y_test=diabetes.target[-20:] ###训练模型
regr=linear_model.LogisticRegression()
regr.fit(diabetes_X_train,diabetes_y_train) ###回归系数
print('Coefficients:\n',regr.coef_) ###均方误差
print('the mean sqare error:%.2f' %np.mean((regr.predict(diabetes_X_test)-diabetes_y_test)**2))
print('Variance score:%.2f' %regr.score(diabetes_X_test,diabetes_y_test))
##散点图
plt.scatter(diabetes_X_test,diabetes_y_test,color='black')
plt.plot(diabetes_X_test,regr.predict(diabetes_X_test),color='blue',linewidth=3)
plt.xticks()
plt.yticks()
b=time.time()
print('the running time is %.2f' %(b-a))
plt.show()

实验结果:

Coefficients:
[ 938.23786125]
Residual sum of squares: 2548.07
Variance score: 0.47
the running time is 0.31


 

Sklearn库例子2:分类——线性回归分类(Line Regression )例子的更多相关文章

  1. 从损失函数优化角度:讨论“线性回归(linear regression)”与”线性分类(linear classification)“的联系与区别

    1. 主要观点 线性模型是线性回归和线性分类的基础 线性回归和线性分类模型的差异主要在于损失函数形式上,我们可以将其看做是线性模型在多维空间中“不同方向”和“不同位置”的两种表现形式 损失函数是一种优 ...

  2. Python机器学习笔记:sklearn库的学习

    网上有很多关于sklearn的学习教程,大部分都是简单的讲清楚某一方面,其实最好的教程就是官方文档. 官方文档地址:https://scikit-learn.org/stable/ (可是官方文档非常 ...

  3. day-10 sklearn库实现SVM支持向量算法

    学习了SVM分类器的简单原理,并调用sklearn库,对40个线性可分点进行训练,并绘制出图形画界面. 一.问题引入 如下图所示,在x,y坐标轴上,我们绘制3个点A(1,1),B(2,0),C(2,3 ...

  4. 复盘一篇讲sklearn库的文章(下)

    skleran-处理流程 获取数据 以用sklearn的内置数据集, 先导入datasets模块. 最经典的iris数据集作为例子. from sklearn import datasets iris ...

  5. 复盘一篇讲sklearn库学习文章(上)

    认识 sklearn 官网地址: https://scikit-learn.gor/stable/ 从2007年发布以来, scikit-learn已成为重要的Python机器学习库, 简称sklea ...

  6. 2.sklearn库中的标准数据集与基本功能

    sklearn库中的标准数据集与基本功能 下面我们详细介绍几个有代表性的数据集: 当然同学们也可以用sklearn机器学习函数来挖掘这些数据,看看可不可以捕捉到一些有趣的想象或者是发现: 波士顿房价数 ...

  7. Python: sklearn库——数据预处理

    Python: sklearn库 —— 数据预处理 数据集转换之预处理数据:      将输入的数据转化成机器学习算法可以使用的数据.包含特征提取和标准化.      原因:数据集的标准化(服从均值为 ...

  8. Python深度学习案例2--新闻分类(多分类问题)

    本节构建一个网络,将路透社新闻划分为46个互斥的主题,也就是46分类 案例2:新闻分类(多分类问题) 1. 加载数据集 from keras.datasets import reuters (trai ...

  9. 分类and分类延展

    1.Category简介 Category,又称为类别&类目&分类,是OC特有语法,在不修改原有类的基础上增加新的方法,一个庞大的类可以多人来分模块开发,有助于团队合作,或者对当前类方 ...

随机推荐

  1. 极客DIY:RFID飞贼打造一款远距离渗透利器

    本文使用最新的渗透工具RFID飞贼(Tastic RFID Thief)和RFID感应破解技术来获取一些拥有安防的建筑物的访问权限. Tastic RFID Thief是一个无声远距离RFID读卡器, ...

  2. java作业2

    (一) 仔细阅读示例: EnumTest.java,运行它,分析运行结果? 你能得到什么结论?你掌握了枚举类型的基本用法了吗? 结论:枚举不属于原始数据类型,它的每个具体值都引用一个特定的对象.相同的 ...

  3. 后台代码对iBatis配置文件中具体的sql语句的调用实现(被封装的增删改查)

    using IBatisNet.Common.Exceptions; using IBatisNet.DataAccess; using IBatisNet.DataAccess.DaoSession ...

  4. ZOJ 3811

    http://acm.zju.edu.cn/onlinejudge/showProblem.do?problemId=5343 网络赛这水题没写过太伤了,赛后写了下1A. 当时钻牛角尖一定要用k次bf ...

  5. /bin/bash^M: bad interpreter: 没有那个文件或目录

  6. turn.js实现翻书效果

    JS插件网 http://www.ijquery.cn/?p=173 描述:Turn.js 是一个轻量级的 (15kb) jQuery/html5 插件用来创建类似书本和杂志翻页效果,支持触摸屏设备. ...

  7. 【题解】【链表】【Leetcode】Add Two Numbers

    You are given two linked lists representing two non-negative numbers. The digits are stored in rever ...

  8. OpenHCI - Data Transfer Types

    There are four data transfer types defined in USB(USB中有4种数据传输类型). Each type is optimized to match th ...

  9. CUDA 标准编程模式

    前言 本文将介绍 CUDA 编程的基本模式,所有 CUDA 程序都基于此模式编写,即使是调用库,库的底层也是这个模式实现的. 模式描述 1. 定义需要在 device 端执行的核函数.( 函数声明前加 ...

  10. (实用篇)php处理单文件、多文件上传代码分享

    php处理  单文件.多文件上传实例代码,供大家参考,具体内容如下 后台处理文件submit_form_process.php <?php /************************** ...