CN104182379A - 一种基于转动惯量的一元线性回归方法 - Google Patents

一种基于转动惯量的一元线性回归方法 Download PDF

Info

Publication number
CN104182379A
CN104182379A CN201410299314.6A CN201410299314A CN104182379A CN 104182379 A CN104182379 A CN 104182379A CN 201410299314 A CN201410299314 A CN 201410299314A CN 104182379 A CN104182379 A CN 104182379A
Authority
CN
China
Prior art keywords
overbar
variable
linear regression
beta
observation
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Pending
Application number
CN201410299314.6A
Other languages
English (en)
Inventor
许蔚蔚
洪亮
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Individual
Original Assignee
Individual
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Individual filed Critical Individual
Priority to CN201410299314.6A priority Critical patent/CN104182379A/zh
Publication of CN104182379A publication Critical patent/CN104182379A/zh
Pending legal-status Critical Current

Links

Landscapes

  • Management, Administration, Business Operations System, And Electronic Commerce (AREA)

Abstract

本发明涉及概率论与数理统计领域,特别是涉及一种基于转动惯量的一元线性回归方法。将变量x,y的成对观测值表示为xy-平面内的样本点集P={(xi,yi)|i∈[1,n]},将P中任一点(xi,yi)视为质量为1的质点,赋予了质量意义的样本点集P相对于任一xy-平面内直线l:y=kx+b的转动惯量为J(k,b),求取J(k,b)的极小值点(k0,b0),将基于观测数据的变量x与y的一元线性回归方程表示为其中,

Description

一种基于转动惯量的一元线性回归方法
技术领域
本发明涉及概率论与数理统计领域,特别是涉及一种基于转动惯量的一元线性回归方法。 
背景技术
线性回归分析是数理统计中最基本的研究方法之一,用以研究变量间的相关关系。在社会经济领域,很多变量间的关系即使在宏观上不是线性的,在微观上仍可近似做线性化处理。另外,有的时候通过对变量进行取对数等预处理,可以将变量间的非线性关系变换为线性关系。目前主流的统计分析、数值计算软件都以矩阵运算为基础。因此,对变量进行高精度的线性回归具有重要的基础作用。 
线性回归根据自变量及因变量的数量可分为一重一元、一重多元、多重多元等几种情况,其中,一重一元线性回归是其中最简单和最基本的问题,简述如下: 
设有变量x,y满足线性关系式y=β01x+ε,其中βi(i=0,1)是常数,ε是随机误差。对各变量进行n次观测,观测值向量为:X=(x1,x2,......,xn)′;Y=(y1,y2,......,yn)′。基于以上观测数据的变量x与y的一元线性回归方程为:一元线性回归方程的矩阵形式为Y=(1,X)B+E,其中,B=(β0,β1)′,E=(ε1,...,εn)′。 
一重一元线性回归最常用的解法是基于最小二乘法的线性回归方法(ordinary least squares regression,OLSR):将y视为因变量,x视为自变量,自变量不视为随机变量,只有因变量视为随机变量;参数矩阵B的极大似然估计为
最小二乘线性回归的结果不具有坐标无关性。所谓坐标无关性指将运算所在坐标系做正交变换(平移或/和旋转)不影响运算的结果。 
社会经济变量中很少有取值不具有随机性的“纯”自变量。由于观察角度、观测仪器、数据定义及归总方法的不同,同一经济现象的观测数据形式上可能有很大差别,但经过某种线性变换甚至简单的坐标变换,数据之间就经常表现出明显的等价性。基于以上理由,希望具有等价关系的数据组的回归结果也相同是很自然的要求,因此,发展具有坐标不变性的线性回归方法是必要的。 
发明内容
本发明提供了一种基于转动惯量的一元线性回归方法,可使回归结果具有坐标无关性。 
为实现上述目的,本发明所采用的技术方案是:基于转动惯量的一元线性回归方法,步骤如下: 
(1)设x和y为具有线性相关关系的两个变量,对这两个变量进行n次观测,x的观测值依 次为x1,x2,......,xn,y的观测值依次为y1,y2,......,yn,x的观测值向量为X=(x1,x2,......,xn)′,y的观测值向量为Y=(y1,y2,......,yn)′; 
(2)将变量x,y的成对观测值表示为xy-平面内的样本点集P={(xi,yi)|i∈[1,n]},i为观测值的序号,将P中任一点(xi,yi)视为质量为1的质点,赋予了质量意义的样本点集P相对于任一xy-平面内直线l:y=kx+b的转动惯量为k为l的斜率,b为l在y轴上的截距,求取J(k,b)的极小值点(k0,b0),其中, b 0 = Y ‾ - k 0 X ‾ , X ‾ = 1 n Σ i = 1 n x i , Y ‾ = 1 n Σ i = 1 n y i , F = Σ i = 1 n ( x i - X ‾ ) ( y i - Y ‾ ) , G = Σ i = 1 n [ ( x i - X ‾ ) 2 - ( y i - Y ‾ ) 2 ] ;
(3)将基于观测数据X和Y的变量x与y的一元线性回归方程表示为其中,  β ^ 1 = k 0 , β ^ 0 = b 0 .
本发明达到的有益效果:使回归结果具有坐标无关性,提高回归精度。 
具体实施方式
本发明的一种基于转动惯量的一元线性回归方法具体步骤如下: 
(1)设x和y为具有线性相关关系的两个变量,对这两个变量进行n次观测,x的观测值依次为x1,x2,.....,xn,y的观测值依次为y1,y2,.....,yn,x的观测值向量为X=(x1,x2,......,xn)′,y的观测值向量为Y=(y1,y2,......,yn)′; 
(2)将变量x,y的成对观测值表示为xy-平面内的样本点集P={(xi,yi)|i∈[1,n]},i为观测值的序号,将P中任一点(xi,yi)视为质量为1的质点,赋予了质量意义的样本点集P相对于任一xy平面内直线l:y=kx+b的转动惯量为k为l的斜率,b为l在y轴上的截距,求取J(k,b)的极小值点(k0,b0),其中, b 0 = Y ‾ - k 0 X ‾ , X ‾ = 1 n Σ i = 1 n x i , Y ‾ = 1 n Σ i = 1 n y i , F = Σ i = 1 n ( x i - X ‾ ) ( y i - Y ‾ ) , G = Σ i = 1 n [ ( x i - X ‾ ) 2 - ( y i - Y ‾ ) 2 ] ;
(3)将基于观测数据X和Y的变量x与y的一元线性回归方程表示为其中,  β ^ 1 = k 0 , β ^ 0 = b 0 .

Claims (1)

1.一种基于转动惯量的一元线性回归方法,其特征在于,步骤如下:
(1)设x和y为具有线性相关关系的两个变量,对这两个变量进行n次观测,x的观测值依次为x1,x2,......,xn,y的观测值依次为y1,y2,......,yn,x的观测值向量为X=(x1,x2,......,xn)′,y的观测值向量为Y=(y1,y2,......,yn)′;
(2)将变量x,y的成对观测值表示为xy-平面内的样本点集P={(xi,yi)|i∈[1,n]},i为观测值的序号,将P中任一点(xi,yi)视为质量为1的质点,赋予了质量意义的样本点集P相对于任一xy-平面内直线l:y=kx+b的转动惯量为k为l的斜率,b为l在y轴上的截距,求取J(k,b)的极小值点(k0,b0),其中, b 0 = Y ‾ - k 0 X ‾ , X ‾ = 1 n Σ i = 1 n x i , Y ‾ = 1 n Σ i = 1 n y i , F = Σ i = 1 n ( x i - X ‾ ) ( y i - Y ‾ ) , G = Σ i = 1 n [ ( x i - X ‾ ) 2 - ( y i - Y ‾ ) 2 ] ;
(3)将基于观测数据X和Y的变量x与y的一元线性回归方程表示为其中, β ^ 1 = k 0 , β ^ 0 = b 0 .
CN201410299314.6A 2014-06-30 2014-06-30 一种基于转动惯量的一元线性回归方法 Pending CN104182379A (zh)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN201410299314.6A CN104182379A (zh) 2014-06-30 2014-06-30 一种基于转动惯量的一元线性回归方法

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201410299314.6A CN104182379A (zh) 2014-06-30 2014-06-30 一种基于转动惯量的一元线性回归方法

Publications (1)

Publication Number Publication Date
CN104182379A true CN104182379A (zh) 2014-12-03

Family

ID=51963440

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201410299314.6A Pending CN104182379A (zh) 2014-06-30 2014-06-30 一种基于转动惯量的一元线性回归方法

Country Status (1)

Country Link
CN (1) CN104182379A (zh)

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US10990718B2 (en) 2017-12-12 2021-04-27 Wipro Limited Method and device for generating physical design parameters of an object
US11685326B2 (en) 2021-11-24 2023-06-27 International Business Machines Corporation Vehicle mass measurement for automated braking

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US10990718B2 (en) 2017-12-12 2021-04-27 Wipro Limited Method and device for generating physical design parameters of an object
US11685326B2 (en) 2021-11-24 2023-06-27 International Business Machines Corporation Vehicle mass measurement for automated braking

Similar Documents

Publication Publication Date Title
Abdullaev et al. Numerical method of solution to loaded nonlocal boundary value problems for ordinary differential equations
Seshadhri et al. Triadic measures on graphs: The power of wedge sampling
Giné et al. Averaging methods of arbitrary order, periodic solutions and integrability
CN103955892A (zh) 一种目标跟踪方法及扩展截断无迹卡尔曼滤波方法、装置
Ji et al. Solving high-order uncertain differential equations via Runge–Kutta method
Karachik et al. Solvability of some Neumann-type boundary value problems for biharmonic equations
CN104102833B (zh) 基于密集区间发现的税务指标归一化与融合计算方法
CN104182379A (zh) 一种基于转动惯量的一元线性回归方法
Kane et al. Determining the number of clusters for a k-means clustering algorithm
CN104281770A (zh) 一种一元线性回归方法
Wang et al. Legendre polynomials method for solving a class of variable order fractional differential equation
CN103926578A (zh) 一种室内环境的线性特征提取方法
Cheng et al. A generic position based method for real root isolation of zero-dimensional polynomial systems
CN104063617A (zh) 一种基于降维超平面的多元线性回归方法
CN107247776A (zh) 一种用于聚类分析中相似度识别的方法
Prata et al. Comparative analysis of robust estimators on nonlinear dynamic data reconciliation
Ramachandran et al. Comparison of arithmetic mean, geometric mean and harmonic mean derivative-based closed Newton Cotes quadrature
Wu et al. Parametric solutions to Sylvester-conjugate matrix equations
Jafarian et al. On Bernstein polynomials method to the system of Abel integral equations
CN105703740A (zh) 基于多层重要性采样的高斯滤波方法和高斯滤波器
Wang et al. Lyapunov-type inequalities for certain higher order fractional differential equations
CN105224806A (zh) 一种获取非晶态物质径向分布的方法
Gao et al. Iterative methods for polynomial equations based on vieta’s theorem
CN103164186A (zh) 一种解决除以零情况的时变倒数计算方法
CN104182380A (zh) 一种基于降维主成分平面的二元线性回归法

Legal Events

Date Code Title Description
C06 Publication
PB01 Publication
C53 Correction of patent for invention or patent application
CB02 Change of applicant information

Address after: 310018 School of management, Zhejiang University of Media and Communications, Xiasha 998, Hangzhou, Zhejiang

Applicant after: Xu Weiwei

Applicant after: Hong Liang

Address before: 100192, Beijing, Haidian District Qinghe clear East Lane 8 building, room 504

Applicant before: Xu Weiwei

Applicant before: Hong Liang

C10 Entry into substantive examination
SE01 Entry into force of request for substantive examination
WD01 Invention patent application deemed withdrawn after publication
WD01 Invention patent application deemed withdrawn after publication

Application publication date: 20141203