Analisis Regresi Komponen Utama Menggunakan NCSS

Masukkan datanya
Analysis > Regression > Prinsipal Componen Analysis
Klik Menu variable
 

 

Pada dependen variable: klik Y
Pada Independnen varaibel: klik X1-x3
Klik Menu Report
Cek List multicolinearity, variance inflation factor, component analysis, PC Coeffisient, write Model dan ANOVA
Klik Run
Berikut Hasilnya:
 
Principal Components Regression Report
Least Squares Multicollinearity Section
Independent          Variance        R-Squared                       
Variable                 Inflation     Vs Other X's         Tolerance
X1                         477.2665               0.9979               0.0021
X2                         485.8581               0.9979               0.0021
X3                           11.7455               0.9149               0.0851
Since some VIF's are greater than 10, multicollinearity is a problem.
 
Variance Inflation Factor Section
PC's                   X1                    X2                    X3
1                   0.2466              0.2511               0.0036
2                   0.2645              0.2513               0.9815
3               477.2665           485.8581             11.7455
 
Components Analysis Section
PC's                   R2                 SSE                   B'B            Ave VIF            Max VIF
1                   0.9905              1.1677               0.4965               0.1671              0.2511
2                   0.9905              1.1674               0.4965               0.4991              0.9815
3                   0.9915              1.1028               1.4905           324.9567           485.8581
 
PC Coefficient Section
Principal                        PC         Individual
Component        Coefficient        R-Squared       Eigenvalue
PC1                          7.6653               0.9905           1.994969
PC2                         -0.0245               0.0000           1.004003
PC3                        -10.8457               0.0010           0.001027
 
Regression Coefficient Section with 1 Component Omitted
                                                                                       Stand'zed
Independent           Regression             Standard           Regression
Variable                  Coefficient              Error                  Coefficient             VIF
Intercept                  0.763326                                                                            
X1                          1.007698                 0.0272776                  0.4945        0.2645
X2                          1.003778                 0.02626337                0.4987        0.2513
X3                          0.568248                 0.2554352                  0.0574        0.9815
Model
0.763326+1.007698*X1+1.003778*X2+0.568248*X3
 
Analysis of Variance Section with 1 Component Omitted
                                          Sum of             Mean                                           Prob
Source                   DF        Squares            Square                F-Ratio             Level
Intercept                  1          9614.223           9614.223
Model                     3          1992.698           664.2327             487.3907           0.000000
Error                       14         19.07968           1.362834
Total(Adjusted)        17         2011.778           118.3399
Mean of Dependent             23.11111
Root Mean Square Error      1.167405
R-Squared                           0.9905
Coefficient of Variation        0.05051271

Search