[D] Approximate the Evidence of bayesian logistic regression

There is an approximation of the evidence of bayesian linear regression in PRML, and maximize the evidence, I would like to know if there is any method on how to approximate the evidence of logistic regression and maximize the evidence.

​

https://preview.redd.it/whs3benpmug91.png?width=1276&format=png&auto=webp&s=dc55d8d7165fc2eaf8ad48946cda3ab6d32905bb

https://preview.redd.it/nlmtmo4qmug91.png?width=784&format=png&auto=webp&s=e940e45c0e290cada4cdcd3ec8ae7de2b29965c6

bayesian evidence logistic regression machinelearning regression