[R] HOw compare 2 models in logistic regression

santoshdvn santoshdvn at gmail.com
Mon Apr 30 10:23:19 CEST 2012


Hi,

I have created 2 models in logistic regression and got the predictor values
are significance on output. Here are 2 summaries of 2 model. HOw can we
compare 2 models by what factor or coefficient and say which model is best

Please help
---------------------------------------------------------------------------------------------------------------
*MODEL 1*

glm(formula = qdataset$Spend_bucket ~ qdataset$Freq +
qdataset$Address_is_res + 
    qdataset$last_update_days_ago, family = binomial)

Coefficients:
                                Estimate Std. Error z value Pr(>|z|)    
(Intercept)                   -1.572e+00  2.344e-01  -6.708 1.98e-11 ***
qdataset$Freq                  8.763e-01  8.229e-02  10.649  < 2e-16 ***
qdataset$Address_is_res       -8.743e-01  2.116e-01  -4.132 3.60e-05 ***
qdataset$last_update_days_ago -4.568e-04  7.625e-05  -5.991 2.09e-09 ***
---
Signif. codes:  0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1 

(Dispersion parameter for binomial family taken to be 1)

    Null deviance: 1228.41  on 988  degrees of freedom
Residual deviance:  957.74  on 985  degrees of freedom
AIC: 965.74

Number of Fisher Scoring iterations: 5
-------------------------------------------------------------------------------------------------------------------------------
*MODEL 2*
glm(formula = qdataset$Spend_bucket ~ qdataset$Freq +
qdataset$Address_is_res, 
    family = binomial)

Deviance Residuals: 
    Min       1Q   Median       3Q      Max  
-2.5798  -0.6428  -0.5894   0.7271   2.2402  

Coefficients:
                        Estimate Std. Error z value Pr(>|z|)    
(Intercept)              -2.6148     0.1732 -15.094  < 2e-16 ***
qdataset$Freq             0.9526     0.0794  11.997  < 2e-16 ***
qdataset$Address_is_res  -0.7622     0.2044  -3.729 0.000192 ***
---
Signif. codes:  0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1 

(Dispersion parameter for binomial family taken to be 1)

    Null deviance: 1228.41  on 988  degrees of freedom
Residual deviance:  995.18  on 986  degrees of freedom
AIC: 1001.2

Number of Fisher Scoring iterations: 5


--------------------------------------------------------------------------------------

Thanks,
Santosh

--
View this message in context: http://r.789695.n4.nabble.com/HOw-compare-2-models-in-logistic-regression-tp4597659.html
Sent from the R help mailing list archive at Nabble.com.



More information about the R-help mailing list