Fechas: 23 y 25 de Enero de 2018 de 12:30 a 14:00
Two of the main tools of applied statistics are logistic regression and Cox’s semi-parametric regression model for lifetime data. Both have the purpose of relating a binary dependent variable - the occurrence of an event of interest (success vs. failure, death vs. survival, etc.) - to a set of covariates (predictors). They are especially important in answering many practical problems in medical statistics, but also in almost every area of application of statistics. In this seminar, I shall present first logistic regression, placing it within the context of the wider class of generalised linear models which also includes Poisson regression and other useful models. Examination of residuals and other diagnostic methods for assessing the adequacy of the model will be discussed. Then, I shall move on to multinomial logistic regression, which is the extension to the situation in which the dependent variable has more than two possible categories. There will also be brief mention of some of the possible models for the special case where these categories are ordered. Finally, Cox’s model will be presented. This is a widely used model for analysing lifetime data (also known as survival analysis), when interest lies not only in whether the event occurs, but when it occurs and how this time is affected by covariates. Participants in the seminar will carry out practical work using R Studio to fit these models to sets of real data.
Si desea recibir un certificado de asistencia al Seminario, debe inscribirse en esta actividad enviando un correo electrónico a email@example.com indicando lo siguiente:
Nombre de la actividad:
Nombre completo del participante:
DNI o pasaporte: