Dr. Giorgio Pedrazzi has received his PhD in Statistical Methodology for Scientific Research from the University of Bologna. He is currently is working as Data Scientist in the Cineca’s HPC Business Unit and he is involved in the development of Machine Learning applications for CINECA customers, micro benchmarks and mini-apps for the European project “Oprecomp” and a novel framework and architecture for the European project “Icarus”. He collaborated to Cineca Data Mining projects developed for the Italian Regulatory Authority for Electricity and Gas and Italian Revenue Agency. He developed Text Mining applications for the Italian Research National Portal and, inside the ASTREA National funded research project, he contributed to develop Information Extraction techniques for the monitoring of the Italian judicial system through the analysis of Court decisions. He was involved in the European project Papyrus: Cultural and historical digital libraries dynamically mined from news archives.
The laboratory work plan addresses the complete process of data analysis, starting with loading data using different approaches and developing models for analyzing and visualizing the results. It will study some programming languages (as R, Python and TensorFlow) with reference to their specific use in High Performance Computing (HPC). In this context it will use libraries for parallel computing (H2O) and libraries for Deep Learning (Keras).
It will cover also software for the creation of graphic Workflow for Data Analytics, as Knime and Orange. The work plan provides for the developing of a project during a Kaggle competition at the end of the course.
The course is coordinated by Giorgio Pedrazzi and will be hold by CINECA staff, the largest Italian computing centre (Dipartimento Super Calcolo, Applicazioni e Innovazione and Laboratorio Big Data e Analytics).