INFORMAZIONI SU
DE MARTINO Maria
 |
Supervisore: Prof. Isola Standard and non-standard statistical inference for discovery-driven research in biological and clinical field |
| The aim of this research project is the use of standard and non-standard biostatistical models for the investigation of biomarkers and their use for diagnosis, prognosis, monitoring, prediction and risk estimation. A crucial point will be the step of validation, essential for the regular application of biomarkers in clinical practice. In doing so the models implemented will be more traditional ones as variations of various type of regression and more innovative ones, as different type of supervised and unsupervised machine learning models. In doing so an interesting aspect will be the comparison of the performance of these different models. Great focus will be put on the study planning and power of the study, both crucial when studying biomarkers and when trying to reduce and avoid bias. These methods will be accomplished through the use and study of STATA, R and Python. There will be many fields of applications: infectious disease, hematology, molecular biology and others. | |