Flexible survival regression modelling: An application to lung cancer data

Cox's proportional hazards regression model has the criticality of not being able to handle time-dependent effects optimally. This dissertation explores a family of more flexible models in survival analysis: additive and multiplicative-additive hazards models. An example of usage and comparison among their performances and prevision capabilities is shown upon application to a dataset concerning patients with NSCLC (non small cells lung cancer), on which blood-biomarkers related to hypoxia, inflammation, immune response and tumour load were measured.