Interpret result of logistic regression in terms of percentage increase or decrease

r
rmachinelearning
machine_learning
logistic_regression

#1

Hi,
Currently I am performing Logistic regression on Attrition Dataset but I want to interpret the result of logistic regression in terms of percentage.I am performing this in R using glm() command and the result which I am getting is in terms of log odds example is given below:
Coefficients:
Estimate Std. Error z value Pr(>|z|)
(Intercept) -0.2831 0.4460 -0.635 0.52561
GenderMale -0.4893 0.2096 -2.335 0.01953 *
Location -0.5556 0.4576 -1.214 0.22464
Grade -1.9760 0.4968 -3.978 6.95e-05

Now I want to convert this Estimate to percentage for that i have used this procedure refered from a document convert log odds to odds to probability and then to percentage.
example : log odds = -0.4893 to odds(o) =exp( -0.4893 ) = 0.6130554 then to probability by (o/1+o) = 0.613/(1+0.613) = 0.380 and then to percentage by (1-p)*100 = 1-0.380=0.62 = 62%

but for log of odds like 3.00878 I am getting percentage as -3953% How to Interpret this result . The document which I am folllowing for this conversion is:
https://rpubs.com/OmaymaS/182726

Kindly give me the correct approach for interpreting result of Logistic regression in percentage.


#2

Probability can not be more than 1. Please refer the image file attached. Open an excel file and put your log odds figure and you will know the probability

2017-07-26_16h02_33


#3

Thanks.