P values in logistic regression and neural networks

I have a question regarding the P values in logistic regression and neural networks.
What is their interpretation in these two methods?
Is there a code or a way to get them in python?


What course are you attending?
What do you mean exactly by “P values”?
Please be specific.

p-value is a measure of correlation between variables, and there are some methods you can use to get p-value for a data such as the well-know Pearson method…
regression or NN algorithms are not methods for that

Thank you for the clarification!
Yes that makes sense now

1 Like

it is my pleasure

Hi Aya, I know the question is old, but I wanted to help with the answer anyway.

A p-value is a concept in hypothesis testing where it is required to have a summary statistics (where we have an “idea” of its probability distribution under different assumptions). Since it is possible to connect a summary statistic with the model’s parameters in logistic regression, then we can perform hypothesis testing around the value of the parameters. However, in neural networks, usually this is not the case, so in most of the cases, we cannot test hypotheses, therefore, we cannot compute p-values.

Here is an example using Python to compute stats to evaluate the logistic regression, including p-values.

Hope this helps.