What is the suitable regression to be used if I have binary independent variables and also the dependent variables ar binary ? i.e. yes/no ?
Details:
I am checking whether the presence/absence of virulence genes is associated with the presence/absence of antimicrobial resistance genes. I coded them all as 0/1. ?
You could use either a logit or a probit model. Both are regression models designed for a binary dependent variable.
You could use logistic regression in this case (Logit link and binary response).
Dear Friend,
SVR is a model of machine learning used for classifying three or more types of something based on numerical values given for variables. This is not our aim here because we just we would like to see how the degree of association between both gene categories.
No, it is all bout setting a correlation between the probability of +/- OR 0/1 which we obtained for the bacterial isolates. The best thing actually is to calculate the correlation. On the other side, regression measures the strength that one or more variable (we called independent or predictor) on the response or we called dependent variable. In the logistic regression, you must have the dependent variable as binary (two results such as live/dead r present/absent) and the independent variables as numbers, which is not our case. Thanks
- Badges
- Science topic
- Similar topics
- Computer Science and Engineering
- Coding