层次分析法(AHP)Python实现

I. 理论基础

  1. 建立层次结构模型

    结构模型包括目标层 O O O、准则层 A = { a 1 , a 2 , … , a n } A=\{a_1,a_2,\ldots,a_n\} A={a1,a2,,an}、方案层 B = { b 1 , b 2 , … , b m } B=\{b_1,b_2,\ldots,b_m\} B={b1,b2,,bm}

  2. 构建判断(成对比较)矩阵

M a t r i x A = [ r 1 , 1 r 1 , 2 … r 1 , n r 2 , 1 r 2 , 2 … r 2 , n ⋮ ⋮ ⋱ ⋮ r n , 1 r n , 2 … r n , n ] Matrix_A = \left[ \begin{matrix} r_{1,1} & r_{1,2} & \ldots & r_{1,n} \\ r_{2,1} & r_{2,2} & \ldots & r_{2,n} \\ \vdots & \vdots & \ddots & \vdots \\ r_{n,1} & r_{n,2} & \ldots & r_{n,n} \\ \end{matrix} \right] MatrixA= r1,1r2,1rn,1r1,2r2,2rn,2r1,nr2,nrn,n

其中, r i , j = 1 r j , i r_{i,j}=\frac{1}{r_{j,i}} ri,j=rj,i1 r i , i = 1 r_{i,i}=1 ri,i=1

  1. 层次单排序与一致性检验

    3.1 计算权重

    (1)若采用均值法计算权重,则首先计算列归一化矩阵:
    N o m a l i z e M a t r i x A = [ n r 1 , 1 n r 1 , 2 … n r 1 , n n r 2 , 1 n r 2 , 2 … n r 2 , n ⋮ ⋮ ⋱ ⋮ n r n , 1 n r n , 2 … n r n , n ] NomalizeMatrix_A = \left[ \begin{matrix} nr_{1,1} & nr_{1,2} & \ldots & nr_{1,n} \\ nr_{2,1} & nr_{2,2} & \ldots & nr_{2,n} \\ \vdots & \vdots & \ddots & \vdots \\ nr_{n,1} & nr_{n,2} & \ldots & nr_{n,n} \\ \end{matrix} \right] NomalizeMatrixA= nr1,1nr2,1nrn,1nr1,2nr2,2nrn,2nr1,nnr2,nnrn,n
    其中, n r i , j = r i , j S j , S j = ∑ i = 1 n r i , j , j = 1 , 2 , … , n nr_{i,j}=\frac{r_{i,j}}{S_j},S_{j}=\sum_{i=1}^n{r_{i,j}},j=1,2,\ldots,n nri,j=Sjri,j,Sj=i=1nri,j,j=1,2,,n

    然后再计算权重,为:
    W = { w 1 , w 2 , … , w n } W=\{w_1,w_2,\ldots,w_n\} W={w1,w2,,wn}
    其中, w i = ∑ j = 1 n n r i , j n , i = 1 , 2 , … , n w_i=\frac{\sum_{j=1}^n{nr_{i,j}}}{n},i=1,2,\ldots,n wi=nj=1nnri,j,i=1,2,,n

    (2)若采用几何法计算权重,则首先按行求积,再计算1/n次幂,计算公式为:
    w i = ∏ j = 1 n r i , j n , i = 1 , 2 , … , n w_{i} = \sqrt[n]{\prod_{j=1}^{n}r_{i,j}},i=1,2,\ldots,n wi=nj=1nri,j ,i=1,2,,n

    然后再对各权重进行归一化处理,计算如下:
    w i ‾ = w i ∑ i = 1 n w i \overline{w_{i}} = \frac{w_i}{\sum_{i=1}^{n}w_i} wi=i=1nwiwi

W = { w 1 ‾ , w 2 ‾ , … , w n ‾ } W=\{\overline{w_{1}},\overline{w_{2}},\ldots,\overline{w_{n}}\} W={w1,w2,,wn}

3.2 计算判断矩阵特征根

(1) 计算加权矩阵

W e i g h t M a t r i x A = [ w r 1 , 1 w r 1 , 2 … w r 1 , n w r 2 , 1 w r 2 , 2 … w r 2 , n ⋮ ⋮ ⋱ ⋮ w r n , 1 w r n , 2 … w r n , n ] WeightMatrix_A = \left[ \begin{matrix} wr_{1,1} & wr_{1,2} & \ldots & wr_{1,n} \\ wr_{2,1} & wr_{2,2} & \ldots & wr_{2,n} \\ \vdots & \vdots & \ddots & \vdots \\ wr_{n,1} & wr_{n,2} & \ldots & wr_{n,n} \\ \end{matrix} \right] WeightMatrixA= wr1,1wr2,1wrn,1wr1,2wr2,2wrn,2wr1,nwr2,nwrn,n
其中, w r i , j = r i , j × w j , j = 1 , 2 , … , n wr_{i,j}=r_{i,j}\times{w_j},j=1,2,\ldots,n wri,j=ri,j×wj,j=1,2,,n

(2) 对加权矩阵按行求和,然后再除以权重,从而得到特征根向量
λ = { ∑ j = 1 n w r 1 , j w 1 , ∑ j = 1 n w r 2 , j w 2 , … , ∑ j = 1 n w r n , j w n } \lambda = \{\frac{\sum_{j=1}^{n}wr_{1,j}}{w_1},\frac{\sum_{j=1}^{n}wr_{2,j}}{w_2},\ldots,\frac{\sum_{j=1}^{n}wr_{n,j}}{w_n}\} λ={w1j=1nwr1,j,w2j=1nwr2,j,,wnj=1nwrn,j}

  1. 计算一致性指标CI值

C I = λ m a x − n n − 1 CI = \frac{\lambda_{max}-n}{n-1} CI=n1λmaxn

其中, C I = 0 CI=0 CI=0表示判断矩阵完全一致, C I CI CI越大,判断矩阵的不一致性程度越严重。

  1. 根据CI和RI值求解CR值,判断其一致性是否通过

    Satty模拟1000次得到的随机一致性指标RI取值如下表所示:

矩阵阶数n12345678910111213
RI000.580.901.121.241.321.411.451.491.511.541.56

C R = C I R I CR = \frac{CI}{RI} CR=RICI
C R < 0.1 CR\lt0.1 CR<0.1,表明判断矩阵 A A A的一致性程度被认为在容忍范围内,此时可用 A A A的特征向量开展权向量计算;若 C R ≥ 0.1 CR\geq0.1 CR0.1,则应考虑对判断矩阵 A A A进行修正。

  1. 计算方案层到指标层的权重

    6.1 建立方案层到指标层的各个方案在每一个指标的两两对比矩阵:
    B k = [ b 1 , 1 k b 1 , 2 k … b 1 , m k b 2 , 1 k b 2 , 2 k … b 2 , m k ⋮ ⋮ ⋱ ⋮ b m , 1 k b m , 2 k … b m , m k ] , k = 1 , 2 , … , m B^k=\left[ \begin{matrix} b_{1,1}^k & b_{1,2}^k & \ldots & b_{1,m}^k \\ b_{2,1}^k & b_{2,2}^k & \ldots & b_{2,m}^k \\ \vdots & \vdots & \ddots & \vdots \\ b_{m,1}^k & b_{m,2}^k & \ldots & b_{m,m}^k \\ \end {matrix} \right] ,k=1,2,\ldots,m Bk= b1,1kb2,1kbm,1kb1,2kb2,2kbm,2kb1,mkb2,mkbm,mk ,k=1,2,,m

    6.2 按照上述操作,计算各个方案的权重
    V k = { v k , 1 , v k , 2 , … , v k , m } , k = 1 , 2 , … , m V_k=\{v_{k,1},v_{k,2},\ldots,v_{k,m}\},k=1,2,\ldots,m Vk={vk,1,vk,2,,vk,m},k=1,2,,m

    最终得到方案到指标的权重矩阵
    V = [ v 1 , 1 v 1 , 2 … v 1 , m v 2 , 1 v 2 , 2 … v 2 , m ⋮ ⋮ ⋱ ⋮ v n , 1 v n , 2 … v n , m ] V=\left[ \begin{matrix} v_{1,1} & v_{1,2} & \ldots & v_{1,m} \\ v_{2,1} & v_{2,2} & \ldots & v_{2,m} \\ \vdots & \vdots & \ddots & \vdots \\ v_{n,1} & v_{n,2} & \ldots & v_{n,m} \\ \end {matrix} \right] V= v1,1v2,1vn,1v1,2v2,2vn,2v1,mv2,mvn,m
    6.3 进行层次总排序一致性判断
    C I i = λ m a x i − n n − 1 , i = 1 , 2 , … , n CI_{i}=\frac{\lambda_{max}^i-n}{n-1},i=1,2,\ldots,n CIi=n1λmaxin,i=1,2,,n

C R = w 1 C I 1 + w 2 C I 2 + … + w n C I n w 1 R I 1 + w 2 R I 2 + … + w n R I n = ∑ i = 1 n w i C I i ∑ i = 1 n w k R I i = C I R I CR = \frac{w_{1}CI_{1}+w_{2}CI_{2}+\ldots+w_{n}CI_{n}}{w_{1}RI_{1}+w_{2}RI_{2}+\ldots+w_{n}RI_{n}}=\frac{\sum_{i=1}^{n}w_{i}CI_{i}}{\sum_{i=1}^{n}w_{k}RI_{i}}=\frac{CI}{RI} CR=w1RI1+w2RI2++wnRInw1CI1+w2CI2++wnCIn=i=1nwkRIii=1nwiCIi=RICI

  1. 计算各个方案的综合评分结果
    S L k = ∑ i = 1 n w i v i , k , k = 1 , 2 , … , m SL_k = \sum_{i=1}^{n}w_{i}v_{i,k},k=1,2,\ldots,m SLk=i=1nwivi,k,k=1,2,,m

参考文献

Thomas L. Saaty. How to make a decision: The Analytic Hierarchy Process. European Journal of Operation Research,19990,48,9-26.

II. 实现流程

AHP实现流程如下:

在这里插入图片描述

III. Python

import numpy as np
from functools import reduce

def ahp_method(dataset, wd = 'm'):
    # 随机一致性指标R.I.(inc_rat)
    inc_rat  = np.array([0, 0, 0, 0.58, 0.9, 1.12, 1.24, 1.32, 1.41, 1.45, 1.49, 1.51, 1.48, 1.56, 1.57, 1.59])
    # 数据深复制
    X        = np.copy(dataset)
    # 生成权重向量
    weights  = np.zeros(X.shape[1])
    # 根据权重计算方法来确定各个权重
    if (wd == 'm' or wd == 'mean'):  #均值
        weights  = np.mean(X/np.sum(X, axis = 0), axis = 1)
    elif (wd == 'g' or wd == 'geometric'):  #几何
        for i in range (0, X.shape[1]):
            weights[i] = reduce( (lambda x, y: x * y), X[i,:])**(1/X.shape[1])
        weights = weights/np.sum(weights)      
    # 计算特征根向量
    vector   = np.sum(X*weights, axis = 1)/weights
    # 获得平均特征根
    lamb_max = np.mean(vector)
    # 计算一致性指标
    cons_ind = (lamb_max - X.shape[1])/(X.shape[1] - 1)
    # 一致性判断
    rc       = cons_ind/inc_rat[X.shape[1]]
    return weights, rc

IV. 测试

案例参考[https://zhuanlan.zhihu.com/p/448412538]

指标层包括:景色( A 1 A_1 A1)、费用( A 2 A_2 A2)、住宿( A 3 A_3 A3)、饮食( A 4 A_4 A4)、交通( A 5 A_5 A5)

方案层包括:苏杭( B 1 B_1 B1)、北戴河( B 2 B_2 B2)、桂林( B 3 B_3 B3)

通过专家打分以及构建以下判断矩阵:

判断矩阵A:

景色费用住宿饮食交通
景色155 1 3 \frac{1}{3} 318
费用 1 5 \frac{1}{5} 511 1 4 \frac{1}{4} 41 1 6 \frac{1}{6} 612
住宿 1 5 \frac{1}{5} 5141 1 5 \frac{1}{5} 513
饮食36516
交通 1 8 \frac{1}{8} 81 1 2 \frac{1}{2} 21 1 3 \frac{1}{3} 31 1 6 \frac{1}{6} 611

得分矩阵分别为:

A 1 = [ 1 2 5 1 2 1 2 1 5 1 2 1 ] , A 2 = [ 1 1 3 1 8 3 1 1 3 8 3 1 ] , A 3 = [ 1 1 3 1 1 3 1 3 1 3 1 ] , A 4 = [ 1 1 3 4 1 3 1 1 1 4 1 1 ] , A 5 = [ 1 1 1 4 1 1 1 4 4 4 1 ] A_{1}=\left[ \begin{matrix} 1 & 2 & 5 \\ \frac{1}{2} & 1 & 2 \\ \frac{1}{5} & \frac{1}{2} & 1 \\ \end{matrix} \right] ,A_{2}=\left[ \begin{matrix} 1 & \frac{1}{3} & \frac{1}{8} \\ 3 & 1 & \frac{1}{3} \\ 8 & 3 & 1 \\ \end{matrix} \right] ,A_{3}=\left[ \begin{matrix} 1 & 1 & 3 \\ 1 & 1 & 3 \\ \frac{1}{3} & \frac{1}{3} & 1 \\ \end{matrix} \right] ,A_{4}=\left[ \begin{matrix} 1 & \frac{1}{3} & 4 \\ \frac{1}{3} & 1 & 1 \\ \frac{1}{4} & 1 & 1 \\ \end{matrix} \right] ,A_{5}=\left[ \begin{matrix} 1 & 1 & \frac{1}{4} \\ 1 & 1 & \frac{1}{4} \\ 4 & 4 & 1 \\ \end{matrix} \right] A1= 121512121521 ,A2= 138311381311 ,A3= 11311131331 ,A4= 131413111411 ,A5= 11411441411

测试代码:

##
A = np.array([
    [1,5,5,1/3,8],
    [1/5,1,1/4,1/6,2],
    [1/5,4,1,1/5,3],
    [3,6,5,1,6],
    [1/8,1/2,1/3,1/6,1],
])
A_1 = np.array([
    [1,2,5],
    [1/2,1,2],
    [1/5,1/2,1],
])
A_2 = np.array([
    [1,1/3,1/8],
    [3,1,1/3],
    [8,3,1],
])
A_3 = np.array([
    [1,1,3],
    [1,1,3],
    [1/3,1/3,1],
])
A_4 = np.array([
    [1,1/3,4],
    [1/3,1,1],
    [1/4,1,1],
])
A_5 = np.array([
    [1,1,1/4],
    [1,1,1/4],
    [4,4,1],
])

# 计算权重和RI
weight_A,RC_A = ahp_method(A)
print('A的权重为:',weight_A)
print('A的RC为:',RC_A)
print()

weight_A1, RC_A1 = ahp_method(A_1)
weight_A2, RC_A2 = ahp_method(A_2)
weight_A3, RC_A3 = ahp_method(A_3)
weight_A4, RC_A4 = ahp_method(A_4)
weight_A5, RC_A5 = ahp_method(A_5)
print('A1的权重为:',weight_A1)
print('A1的RC为:',RC_A1)
print('A2的权重为:',weight_A2)
print('A2的RC为:',RC_A2)
print('A3的权重为:',weight_A3)
print('A3的RC为:',RC_A3)
print('A4的权重为:',weight_A4)
print('A4的RC为:',RC_A4)
print('A5的权重为:',weight_A5)
print('A5的RC为:',RC_A5)

## 构建V矩阵
V = np.vstack((weight_A1,weight_A2,weight_A3,weight_A4,weight_A5))
print(V)

scores = np.matmul(weight_A,V)
print('综合评分为:',score

测试结果:

A的权重为: [0.30685018 0.06313468 0.12601939 0.45879775 0.04519801]
A的RC为: 0.09521589663079191

A1的权重为: [0.59488796 0.27661064 0.1285014 ]
A1的RC为: 0.004774741975367846
A2的权重为: [0.08199023 0.23644689 0.68156288]
A2的RC为: 0.0013293778439305783
A3的权重为: [0.42857143 0.42857143 0.14285714]
A3的RC为: 0.0
A4的权重为: [0.48036759 0.26858814 0.25104428]
A4的RC为: -0.18508347602743516
A5的权重为: [0.16666667 0.16666667 0.66666667]
A5的RC为: 0.0
[[0.59488796 0.27661064 0.1285014 ]
 [0.08199023 0.23644689 0.68156288]
 [0.42857143 0.42857143 0.14285714]
 [0.48036759 0.26858814 0.25104428]
 [0.16666667 0.16666667 0.66666667]]
综合评分为: [0.46965078 0.28457497 0.24577426]
评论
添加红包

请填写红包祝福语或标题

红包个数最小为10个

红包金额最低5元

当前余额3.43前往充值 >
需支付:10.00
成就一亿技术人!
领取后你会自动成为博主和红包主的粉丝 规则
hope_wisdom
发出的红包
实付
使用余额支付
点击重新获取
扫码支付
钱包余额 0

抵扣说明:

1.余额是钱包充值的虚拟货币,按照1:1的比例进行支付金额的抵扣。
2.余额无法直接购买下载,可以购买VIP、付费专栏及课程。

余额充值