Professor Porcu’s main research interests lie within Statistical and Machine Learning, Data Science, and Spatial Statistics. He has published over 160 peer reviewed papers in top journals in Statistics, Machine Learning, and Mathematics. Applications of his research have involved climate change, natural and anthropogenic catastrophes, weather forecast as well as spatial criminology. More recently, he has been increasingly involved in research related to Health Data Science, with broad applications to computational genetics, genomics, and aging.

Seminar Overview:

Data Science has seen an unprecedented growth in the last 30 years. Nowadays, there is a proliferation of professionals who claim to be a data scientist, and the job applications resort to using buzz words without much depth. This talk revisits the literature and provides a landscape where even the basic definition of Data Science remains unclear. The insatiable appetite for data and for "getting things done" have led to contradictions, inconsistencies, and paradoxical claims. This talk argues that some comprehensive theory is necessary to guarantee the consistency of a system of knowledge. One omnipresent concern is the impact on the reproducibility crisis. The talk offers recipes to avoid false discoveries and provides some suggestions for future research.