Giorgio Corani

• Associate Professor of Data Science and Applied Statistics at IDSIA (USI - SUPSI, Switzerland).

• Member of the IPG group.

Research topics

• Probabilistic forecasting and reconciliation

• Bayesian Data Analysis and applied statistics

• Probabilistic Graphical Models

Recent seminars

Advances in probabilistic reconciliation, Monash University, Melbourne, 2025.

Probabilistic reconciliation via conditioning, invited talk at the 2023 IIF Workshop on Forecast Reconciliation  

Publications

• My list of publications 

• Google Scholar profile

• Orcid profile

• Top 2% scientist (2022 , 2023)  (Stanford's database) 

Conferences organization

I have co-organized, co-chaired and co-edited the proceedings of:

• PGM 2016  (The Eighth International Conference on Probabilistic Graphical Models) 

• ISIPTA 2017 (The Tenth International Symposium on Imprecise Probability: Theories and Applications) 

Program committees  

• AISTATS (2020, 2024, 2025)

• IJCAI (2015, 2016, 2018, 2019, 2021, 2022, 2024)

• UAI (2016, 2018, 2019, 2020, 2021, 2025)

• AAAI (2018, 2019, 2020, 2021, 2022)

• NeurIPS (2018, 2019, 2020, 2024)

• ECML (2023) 

• ECAI (2014, 2016, 2018)

• DMIN (2011 , 2012 , 2013 , 2014)

• ISIPTA (2009, 2011, 2013, 2015, 2017, 2019)

• PGM (2014, 2016, 2018, 2020)

Teaching

Uncertain Reasoning and Data Mining (2010 --2019)

Master of Information Technology, USI (co-teacher).

Analysis of Sequential Data (2018 - 2022)

Master MSE of Data Science, Zurich (co-teacher).

Applied Statistics (since 2010)

Bachelor of Business Engineering, SUPSI. 

Bayesian Data Analysis and Probabilistic Programming (since 2022)

Bachelor of Data Science and Artificial Intelligence, SUPSI.

Software packages

bayesRecon: Probabilistic Reconciliation via Conditioning

Probabilistic reconciliation of base forecast for Gaussian and count variables.

GP forecast: automatic forecasting with Gaussian Processes

Based on our ECML-PKDD 2021 paper; interface for Python and R.

Bayesian hypothesis testing in machine learning

Slides and code of our tutorial at ECML-PKDD 2016.

G. Corani, A. Benavoli, J. Demsar

Naive Credal Classifier

JNCC2 is the implementation of the naive credal classifier (JMLR 2008)

Credal Model Averaging 

The R implementation of credal model averaging for logistic regression is due to A. Mignatti. 

The software and the data set used are available here.