The talk starts at 12:15.

Please note that due to COVID-19, the participants can watch the streamed talk on Teams with a link (below).

Speaker: Simon Boge Brant

Location: 8th floor math building and Zoom. See link below.

Title: A copula-based extension of logistic regression

Abstract: In machine learning, classification algorithms are often described in terms of the dichotomy of being either discriminative or generative. The former is modelling the conditional expectation through regression, for instance logistic regression, and the latter is to model the distribution of the covariates conditional on the outcome, for instance a so-called naîve Bayes model. In this talk, the topic of which is joint work with my supervisor Ingrid Hobæk Haff and Haakon Bakka, I will describe how one can find a naive Bayes model that is equivalent to a fitted logistic regression model, in terms of the regression function. In addition, I will discuss how this connection can be used as inspiration for an extension of the logistic regression model, how we have chosen to do this, and some of the practical challenges involved in constructing this extension.

Join Zoom Meeting:
https://uio.zoom.us/j/63199595088?pwd=Y05GYkJwR1dCUHFhUnlGMFpsdGc3UT09

Meeting ID: 631 9959 5088
Passcode: 302551

Welcome!
Best regards,
Thea Roksvåg and Lars Henry Berge Olsen.